Insulin Glargine, injection, 100 units per mL, 10 mL vials and 3 mL x 5 cartridges, Lantus® , July 2005

Public Summary Document for Insulin Glargine, injection, 100 units per mL, 10 mL vials and 3 mL x 5 cartridges, Lantus® , July 2005.

Page last updated: 24 October 2005

Public Summary Document
Product: Insulin Glargine, injection, 100 units per mL, 10 mL vials and 3 mL x 5 cartridges, Lantus®
Sponsor: sanofi-aventis Group
Date of PBAC Consideration: July 2005

1. Purpose of Application

This application sought listing of Insulin Glargine on the Pharmaceutical Benefits Scheme (PBS) as an unrestricted benefit.

2. Background

This is the fourth submission to the Pharmaceutical Benefits Advisory Committee (PBAC) for Insulin Glargine.

At the November 2004 meeting, the PBAC rejected the third submission. This submission sought an authority required listing restricting subsidised access through the PBS to:
- the treatment of type 1 and type 2 diabetic patients who have inadequate blood glucose control or recurrent unexplained severe hypoglycaemia;
- the treatment of insulin dependent diabetes in children under the age of 18; and
- those who are part of the current Commonwealth special access program supplying pharmaceutical benefits to remote Aboriginal Health Services.

The basis for the rejection of the November 2004 re-submission was the uncertain but modest extent of clinical benefit and unfavourable, albeit uncertain, cost effectiveness. The PBAC recognised the potential importance of an insulin preparation that can reduce the incidence of hypoglycaemia, whilst maintaining good glycaemic control. However, the Committee considered that the evidence presented did not demonstrate significant differences between Insulin Glargine and insulin NPH in this regard. Thus any recommendation for listing in a broader population (the PBAC’s preference) could not be at the price requested.

3. Registration Status

Insulin Glargine, 100 units per mL, 10 mL vials and 3 mL cartridges (Optipen) were registered by the Therapeutic Goods Administration (TGA) on 19 February 2001 for the treatment of type I and type II diabetes mellitus in adults who require insulin for the control of hyperglycaemia.

The TGA approved indication was extended on 6 March 2002 to include the treatment of type 1 diabetes mellitus in children aged 6-15 years.

Insulin Glargine cartridges, 3mL, in a disposable pen injector presentation, (OptiSet), was registered by the TGA on 5 June 2003.

4. Listing Requested and PBAC’s View

The submission sought listing as an unrestricted benefit.

At the November 2004 PBAC meeting, the Committee expressed a preference for a listing in a broader population when considering that submission’s proposed “authority required” listing which restricted listing to a limited subgroup of the insulin-dependent diabetic population. The Committee however indicated this could not be at the price requested at that time.

The Committee noted the current submission requested an unrestricted listing at a lower price for Insulin Glargine compared to the previous submission. This is to be achieved through a risk sharing arrangement.

5. Clinical Place for the Proposed Therapy

The management of insulin dependent diabetes may be complicated by the trade-off between achieving better glycaemic control, as measured by a reduction in HbA1c levels, and a greater risk of hypoglycaemia resulting from a temporary relative oversupply of insulin seen with existing intermediate acting insulins. Insulin Glargine provides a smooth, peakless predictable time concentration profile and a prolonged duration of action, allowing once daily dosing to meet a patient’s basal insulin needs.

6. Comparator

The submission nominated insulin NPH as the comparator. PBAC had previously accepted this as appropriate.

7. Clinical Trials

The submission included information from 19 randomised clinical trials of Insulin Glargine compared with insulin NPH. The submission also presented a negative binomial regression analysis of the hypoglycaemic event rate data from the sub-set of these 19 trials for which individual patient data were available.

Evidence from survey-based studies in support of the impact of hypoglycaemia and the fear of hypoglycaemia was also presented.

The studies that were published or had been presented at conferences at the time of submission are detailed in the table below:

Trial/first author

Protocol title

Publication citation

3004/ Ratner RE 28-week multicentre controlled randomised open clinical trial comparing HOE901 with NPH human insulin in subjects with Type 1 diabetes. Diabetes Care May 2000; 23(5): 639-43.
3005/ Raskin P 16-week multicentre controlled randomised open-label clinical trial comparing HOE901 insulin and NPH human insulin in subjects with Type 1 diabetes treated with insulin Lispro. Diabetes Care 2000; 23(11): 1666-71.
3002/
1. Yki-Jarvinen
2. Benedetti M
52-week multicentre controlled randomised open clinical trial comparing HOE 901 with NPH human insulin subjects with Type 2 diabetes. 1. Diabetes Care 2000; 23(8) 1130-36
2. Hormone Metabolism Research (2003); 35: 189-196
3006/ Rosenstock J 28-wk multicentre controlled randomised open clinical trial comparing HOE 901 with NPH human insulin in subjects with Type 2 diabetes. Diabetes Care 2001; 24(4) 631-36.
4010/ Fulcher G 30-week multicentre controlled randomised single blind clinical trial comparing Insulin Glargine and insulin Lispro with NPH human insulin and Insulin Lispro in subjects with Type 1 diabetes. Diabetologia (2002): 43 (suppl) European Association for the Study of Diabetes (EASD) September 2002 Abstract 801.
4001/ Fritsche A 28-week open controlled randomised multinational multicentre clinical study to investigate the efficacy and safety of different combination therapies, HOE901 insulin analogue (once daily, at bedtime or in the morning) plus glimepiride and NPH basal insulin (once daily, at bedtime) plus glimepiride, in Type 2 diabetes mellitus patients who fail good metabolic control with oral antidiabetic drugs. Diabetologia (2002); 45 (Suppl 2); 38th annual meeting of the European Association for the Study of Diabetes (EASD) Abs: 149.
4002/ Riddle Target glycaemic control and the incidence of symptomatic nocturnal hypoglycaemia in insulin naïve subjects with Type 2 diabetes on oral hypoglycaemic agent(s) and treated with Insulin Glargine or NPH human insulin. 62nd Sci Sess Am Diabetes Assoc (ADA), June 2002, Diabetes 2002; 51 (suppl 2): A113, abs 457-P.
4006/ Ashwell S 32-week, five centre, two-way cross-over study in Type 1 diabetes comparing Insulin Glargine and insulin Lispro with NPH and unmodified human insulin. ADA: conference poster 784
4012/Presented as: Chang Yu Pan 24-week, open, controlled, randomised, multinational, multicentre, phase IIIb clinical study to investigate the efficacy and safety of Lantus insulin analogue (once daily at bedtime) plus Amaryl (glimepiride) and NPH basal insulin (once daily at bedtime) plus Amaryl in 440 subjects with type 2 diabetes mellitus who fail good metabolic control with oral antidiabetic drugs. Presented at ADA 2004
Connolly M The epidemiology and impact of hypoglycaemia in an Australian population: evaluation by postal survey, Australian Diabetes Society Annual Meeting 2002. Poster.
Dixon S The interaction between frequency of hypoglycaemia, single and multiple vascular complications and health utility (poster). 38th Annual Meeting of the European Association for the Study of Diabetes. 2003 Abstract #304.
Hershon K Lower fasting blood glucose (FBG) and less symptomatic hypoglycaemia with QD Lantus (LANTUS ®) compared to BID NPH in subjects with Type 1 diabetes. Diabetes (2001); 50(2): A116-A117. 61st Sci Sess Am Diab Association, Philadelphia (June 2001); Abstract 466-P
Johnson R (A) International Differences in Acceptable Trade-Offs Between Glucose Control And Hypoglycemia: Results of a Physician Survey in Five Countries ADA 63rd Scientific Sessions, New Orleans, 2003, and also 18th IDF Congress, Paris, August 2003, Diabetes and Metabolism 2003 29(Spec No 2): 4S408 (Abstract 2924), Diabetes Care 2003; 52 (Suppl 1): A263 (Abstract 1133).
Johnson R (B) Trade-offs between glucose control and hypoglycaemia in different patient types: results of a 5-country physician survey. Diabetes Care 2003; 52 (Suppl 1): A264 (Abstract 1134).
Pieber TR Efficacy and safety of HOE 901 versus NPH Insulin in patients with Type 1 diabetes. Diabetes Care (2000); 23: 157-162.
Schober E Comparative trial between Insulin Glargine and NPH Insulin in children and adolescents with Type 1 Diabetes Mellitus. Journal of Pediatric Endocrinology and Metabolism (2002); 15(4): 369-76.
Matthews DR A new long-acting insulin (HOE 901) demonstrates less nocturnal hypoglycaemia when compared with protamine insulin in a clinical trial. Diabetologia 1998; 41(1):A245. 34th Ann Mtg Eur Assoc Study Diabetes, Barcelona (Sep 1998). Abstract: 948.
Porcellati F Glargine vs NPH as Basal Insulin in Intensive Treatment of T1DM Given Lispro at Meals: One Year Comparison. ADA Abstract, 2002.
Rossetti P. Intensive replacement of basal insulin in patients with Type 1 diabetes given rapid-acting insulin analog at mealtime. A 3-month comparison between administration of NPH insulin four times daily and glargine insulin at dinner or bedtime. Diabetes Care, Volume 26, Number 5, May 2003, pp 1490-1496
Rosenstock J, Basal Insulin Glargine (HOE 901) vs. NPH Insulin in Patients with Type 1 Diabetes on Multiple Daily Insulin Regimens. Diabetes Care, 2000 23(8): 1137-42.
Yki-Jarvinen, H The LANMET Study. No: 2181 – P0. ADA 2004 – Study 6001

 

8. Results of Trials

The outcomes used in the regression analyses were the relative rate reductions in “symptomatic”, “major” and “severe” hypoglycaemic events per patient for Insulin Glargine compared with insulin NPH expressed in percentage terms and standardised for differential durations of follow-up. A severe hypoglycaemic episode was defined as one with symptoms consistent with hypoglycaemia and requiring the assistance of another person and blood glucose < 2.8mmol/L or prompt recovery with oral carbohydrate, intravenous glucose or glucagon.

Based on the negative binomial regression there was a statistically significant reduction in symptomatic hypoglycaemia (8.6%) and major hypoglycaemia (27.9%) relative to insulin NPH. Although no significant reduction in “severe” hypoglycaemic events was seen in the analysis, the PBAC considered that this result could be an artefact. The reductions for all types of hypoglycaemic events were trending in the same direction and all were congruent, however the numbers for severe events were small. Notwithstanding, the absolute sizes of the reductions in the types of hypoglycaemic events with Insulin Glargine were small.

The PBAC noted that given the small reductions in the actual number of events, the clinical importance of these reductions is questionable and it does not totally remove the risk of hypoglycaemic events.

No new toxicity data were presented in the submission. Based on the data presented in the second submission (June 2003) and the original submission, there appears to be a similarity in the occurrence of adverse events (other than hypoglycaemic events) with Insulin Glargine and NPH insulin.

The PBAC had a number of other concerns with the evidence presented in relation to hypoglycaemic events (see Recommendations and Reasons).

9. Clinical Claim

The submission stated that Insulin Glargine is no worse than insulin NPH in terms of effectiveness (as measured by the primary outcome of HbA1c reduction), but claimed that it is less toxic in terms of rates of hypoglycaemic events.

The PBAC reaffirmed its previously expressed view that hypoglycaemia is a significant issue for insulin treated diabetic patients. The PBAC recognises that there are patients who will potentially benefit from the advantage in hypoglycaemic rates with Insulin Glargine over insulin NPH. However the data presented led PBAC to conclude that there is uncertain overall benefit in the insulin-dependent diabetic population as a whole, as would become eligible under the proposed PBS listing restriction (see Recommendations and Reasons).

10. Economic Analysis

An updated preliminary (trial-based) economic evaluation was presented. The submission estimated the trial-based incremental costs per extra “symptomatic”, “major” or “severe” hypoglycaemic event avoided for Type I and Type II diabetes separately and for vials and cartridges separately, giving a total of 12 separate incremental cost calculations. The resulting incremental costs were <$15,000. The lower incremental costs were associated with the more common “symptomatic” events, and the higher incremental costs with the less common “severe” events.

A number of problems with this analysis were identified during the evaluation, and the PBAC considered that the trial-based incremental costs per extra hypoglycaemic event avoided could be higher than estimated in the submission.

An updated modelled economic evaluation was presented. The base case modelled incremental discounted costs per extra discounted quality adjusted life year (QALY) gained fell in the ranges <$15,000 for Type I diabetes and $15, 000 – $45,000 for Type II diabetes.

The PBAC had a number of major concerns with the economic model presented. As a result of these concerns the Committee concluded that the incremental cost per extra quality-adjusted life-year gained is highly uncertain and under one scenario could be >$200,000 in Type I diabetes and higher in Type II diabetes (see also Recommendation and Reasons).

11. Estimated PBS Usage and Financial Implications

The submission estimated the annual number of patients to be in the range 50,000 – 100,000 for Type 1 diabetes and 50,000 – 100,000 for Type 2 diabetes by the forth year after listing.

The submission estimated additional direct costs to the PBS in the range of $10 - 30M in Year 1, increasing to > $60M in Year 4.

The PBAC considered the estimates presented in the submission to be low. The main areas of uncertainty are in the number of type II patients who are likely to take up Insulin Glargine.

12. Recommendation and Reasons

The PBAC reaffirmed its previously expressed view that hypoglycaemia is a significant issue for insulin treated diabetic patients. The PBAC recognises that there are patients who will potentially benefit from the advantage in hypoglycaemic rates with Insulin Glargine over insulin NPH.

The Committee noted the sponsor had requested an unrestricted listing at a lower price for Insulin Glargine compared to the previous submission. This is to be achieved through a risk sharing arrangement.

The PBAC noted that the regression analysis presented in the submission showed significant reductions in “symptomatic” and “major” hypoglycaemic events per patient for Insulin Glargine compared with insulin NPH. Although no significant reduction in “severe” hypoglycaemic events was seen in the analysis, this result could be an artefact. The reductions for all types of hypoglycaemic events were trending in the same direction and all were congruent, however the numbers for severe events are small. Notwithstanding, the absolute sizes of the reductions in the types of hypoglycaemic events with Insulin Glargine were small and it does not totally remove the risk of hypoglycaemic events. When trials for the two types of diabetes are considered separately, the rates of events for Type II diabetes are smaller, and the corresponding reductions in the numbers of events per patient per year are also smaller.

The meta-analysis included in the current submission uses the three categories of “symptomatic”, “major” and “severe” hypoglycaemia. These categories are poorly defined, and differ across trials, reducing the robustness of the case for the meta-analysis and the cost-utility analyses that rely on these categories. In addition, although noting that the sponsor had access to patient level data for these trials only, the fact remains that only a sub-set of the nineteen trials in the submission were included in the meta-analysis. Trial-level meta-analyses of the total trial set and trial sub-set would allow an assessment of whether the submission is based on a biased sample of trials.

The PBAC noted that an implication of the statement by the sponsor (used in addressing an issue regarding the number of injections experienced by patients which was raised during the evaluation) that“[t]he majority of people using NPH basal insulin does so in a twice-daily (BD) dose” is that this majority would obtain no demonstrated therapeutic advantage by the substitution of Insulin Glargine.

The absolute differences in hypoglycaemic event rates used in the modelled economic evaluation were higher than those observed in the clinical trials, and were considered unlikely to reflect real life by the Committee. For example, severe hypoglycaemic events in Type 1 patients treated with insulin NPH and Insulin Glargine were estimated to occur approximately 3 and 2 times per year, respectively (compared to less than once per year from the trial data), thus the largest discrepancies between the modelled rates and the trial rates impacted on the largest but most uncertain relative risk reductions with the greatest claimed utility impacts.
The Connelly survey upon which these event rates were based might not be representative.

At its hearing, the sponsor asserted that “real life” event rates for hypoglycaemia were likely to be higher than those observed in clinical trials as non-compliant patients are generally excluded from these trials. The Committee did not accept this argument, noting that patients who are not compliant with their insulin treatment could also experience fewer rather than more hypoglycaemic events and that the submission emphasises the aggressive insulin titration intent of the more recently conducted trials which tended to increase the hypoglycaemic event rates. The use of the higher event rate differences in the model favoured Insulin Glargine. Even if the Committee accepted the claimed reduction in “severe” events from three to two per year, such a reduction may not be clinically relevant, with two “severe” hypoglycaemic events per year still being important.

A related concern is the discrepancy introduced by adjusting hypoglycaemic events rates in the model whilst retaining the trial-based doses in Type I and Type II diabetes to estimate drug costs in the model, which undermines the internally valid argument based on the trials that differing doses of the two insulins (together with doses of concomitant insulin lispro) also influence the rate of hypoglycaemia.

A repeated assertion of expert opinion in the submission is that fear of hypoglycaemia results in patients self-titrating their insulin dose downwards, thus compromising their glycaemic control. Although plausible, this assumption is not supported by any presented evidence.
The primary basis of the submission is to relate differences in hypoglycaemic rates to differences in quality-adjusted life-years gained. The utility values used in the economic model were considered poorly justified. In particular, the disutility values used to estimate the fear of hypoglycaemia are important drivers of the modelled evaluation, but are of doubtful validity, and are based on divergent studies which do not adequately address ‘fear’ of hypoglycaemia. There are large increases in estimated disutilities for the various types of hypoglycaemia compared to the previous submission. As an example of the doubts about the validity of the estimated disutility associated with fear of hypoglycaemia, the Connolly study is cited to support the disutility of fear associated with each severe hypoglycaemic event, whereas what appears to be estimated is confounded by the concurrent impact of long-term complications and co-morbidities rather than being attributable to the loss of utility associated with a particular hypoglycaemic event. As a result the incremental cost per extra quality-adjusted life-year gained is highly uncertain and could be > $200,000 in Type 1 diabetes and higher in Type II diabetes in one scenario.

The Committee accepted that the model does not irreversibly cumulate fear over successive hypoglycaemic events over the entire duration of the model. However, the PBAC also noted that the submission had not justified its quantitative assumptions in relation to accumulation of disutility of fear in each cycle period, including by not explaining how this period refers to the different periods of time surveyed in the various studies cited to support the existence of a relationship (eg. 3 months with Dixon et al, 2003 and 6 months with the Connolly survey).

This is also an important possible source of structural uncertainty in the model given the extent to which this disutility drives the model as evidenced by the sensitivity analyses presented in the evaluation report.

Thus, although the PBAC accepted that hypoglycaemic events have an impact on fear of future events and therefore an impact on disutility, the specific nature of any reduction in fear by substituting Insulin Glargine in place of insulin NPH was not adequately demonstrated and the associated estimates to quantify the extent of the difference in the extent of overall fear-related disutility were considered to be particularly poorly justified.

The PBAC did not accept other assumptions in the economic model. If Insulin Glargine does reduce hypoglycaemic event rates to differing degrees across Type I and Type II diabetes, then applying a combined rate to separate Type I and Type II models will be in error to the extent determined by the relative weight of Type I and Type II trials in the meta-regression.

Thus, although the PBAC continues to agree that there are patients who will potentially benefit from the advantage in hypoglycaemic rates with Insulin Glargine over insulin NPH, the submission was rejected on the grounds of uncertain overall benefit in the insulin-dependent diabetic population as a whole and of uncertain and unacceptably high cost-effectiveness at the requested price, which remains significantly higher than the current PBS price for insulin NPH.

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

Sanofi-aventis considers that the PBAC rejected significant elements of the data presented in its submission without appropriate justification. These are discussed in detail on the sponsor’s website www.sanofi-aventis.com.au. The particular issues that sanofi-aventis disputes are:

1. The PBAC's conclusion that the impact of the reduction in severe hypoglycaemia is of questionable clinical importance
2. The PBAC's concerns regarding the utility data used in the economic model
3. The combining of the Type I and Type II diabetes data rather than reporting these separately.

While sanofi-aventis is disappointed by the PBAC rejection, importantly, the PBAC has recently held a stakeholder’s meeting to discuss a way forward for Insulin Glargine and the company is planning to resubmit to the March 2006 PBAC meeting.