Lantus, the reference insulin glargine used for the treatment of diabetes, lost its patent protection in 2014, opening the market to biosimilar competitors.
First, to analyze the adoption rates of insulin glargine biosimilars in primary care in England and estimate the savings realized and missed, since an insulin glargine biosimilar was first used, and second, to assess potential variations in adoption rates across Clinical Commissioning Groups (CCGs).
Data sets capturing information on all insulin glargine items prescribed by all general practitioners up to December 2018 were used. Total costs of insulin glargine and uptake rates of biosimilars were calculated. The real-world budget impact was estimated assuming the cost of reference insulin glargine for all items and comparing the total costs in this scenario with the total costs in the real world. The missed savings were estimated assuming the cost of biosimilars for all insulin glargine items. Choropleth maps were generated to assess potential variations in uptake across CCGs.
Insulin glargine biosimilars generated savings of £900,000 between October 2015 (time of first prescription) and December 2018. The missed savings amounted to £25.6 million in this period, indicating that only 3.42% of the potential savings were achieved. The analyses demonstrated a large level of variation in the uptake of insulin glargine biosimilars across CCGs, with market shares ranging from 0 to 53.3% (December 2018).
These results may encourage decision makers in England to promote the use of best-value treatments in primary care and to reevaluate variation across CCGs.
Introduction
Biosimilars, as defined by the European Medicines Agency (EMA), are medicines considered to be highly similar to another medicine already marketed in the European Union (i.e., the reference product). Owing to the natural variability associated with the production of biological medicines, the EMA acknowledges that minor differences can exist between the biosimilar and its reference product but that these are not meaningful in terms of quality, safety, and efficacy according to the EMA (1). Prices of biosimilars are typically lower than those of the corresponding reference products, and National Health Service (NHS) England estimated that potential savings with biosimilar adoption could reach £300 million per year by 2021 (i.e., equal to 1.8% of the total expenditure on medicines in 2015–2016) (2). For this reason, NHS England put in place a Commissioning Framework in 2017 to optimize the use of biological medicines and promote the use of biosimilars when clinically appropriate (2). This framework emphasized the potential for improvement in specialized services, focusing on savings that could be realized from a wide adoption of adalimumab, etanercept, infliximab, rituximab, and trastuzumab biosimilars (2). Despite its focus on secondary-care medicines, the potential savings from the adoption of biosimilars in primary care, such as those resulting from the uptake of insulin glargine biosimilars, should not be underestimated.
The patent on the Lantus, the reference insulin glargine for the treatment of diabetes in adults, adolescents, and children aged 2 years and older, expired in 2014 in Europe and the U.S. (3,4). Since then, several companies have launched insulin glargine biosimilars, including Abasaglar (approved by the EMA in September 2014) and Semglee (approved by the EMA in January 2018) (3). After the approval of Abasaglar, the Association of British Clinician Diabetologists (ABCD) released a statement acknowledging the potential savings that insulin glargine biosimilars could bring to the NHS without compromising patients’ safety or efficacy outcomes and supported the use of these treatments for newly diagnosed patients (5). The ABCD considered, however, that switching patients on treatment should not be done automatically (neither at the hospital, general practitioner [GP] practice, nor pharmacy level) and should only be conducted under properly trained clinical teams (5). In line with this and NHS England’s Commissioning Framework, a number of regional and local authorities have put in place policies to promote the use of biosimilars, including insulin glargine. For instance, the South Staffordshire Area Prescribing Group, which includes several Clinical Commissioning Groups (CCGs), published a statement underlining that it will “initiate new patients requiring insulin glargine on biosimilar Abasaglar” (6). Also, several CCGs have issued clear guidance on how to switch patients on Lantus to Abasaglar and even calculated the savings that could be realized from this change (7–11).
In general, however, the CCG Biosimilar National Questionnaire results suggest that CCGs put more emphasis on the development of uptake plans for biosimilars used in secondary care; for instance, 98% and 96% of CCGs confirmed having developed plans for commissioning infliximab and etanercept biosimilars for new patients, with only 3% of CCGs indicating that similar strategies had been put in place for insulin glargine biosimilars (12). Therefore, a considerable amount of potential savings may have been missed, particularly in primary care, the health care setting where diabetes patients are regularly treated and monitored (13). For this reason, the main objective of our analysis was to estimate the real-world adoption rates and budget impact, or savings, of insulin glargine biosimilars in primary care in England, as well as the savings that may have been missed since their launch. In addition, we assessed whether insulin glargine biosimilar adoption rates varied across CCGs and attempted to identify whether different prescribing policies may have led to any potential differences.
Research Design and Methods
Data
In the interest of good governance and public accountability, the U.K. Government has committed to increase transparency in the public sector in recent years, including making publicly available prescribing data at the GP practice level (14). On 21 May 2019, we extracted the data on each individual treatment (identified by a unique British National Formulary code) prescribed by each GP practice in England and dispensed in the community in the U.K. Prescribing information data sets were available monthly, recording data starting from August 2010 until December 2018 (at the time of data collection). We filtered these data sets to capture prescriptions of insulin glargine prescribed by brand name and excluded those where insulin glargine was prescribed generically; the products and presentations included in the analyses are described in Table 1. The data sets captured a small number of Lantus OptiSet and Lantus OptiClik prescriptions, but these were excluded from the analyses due to the manufacturer’s decision to discontinue their commercialization following the advice from the National Patient Safety Agency in 2010 (15). Toujeo DoubleStar and Semglee (an insulin glargine biosimilar) were not captured in our data set due to the time limits (i.e., December 2018).
Products included in the analysis
| Product | Presentation | Reference or biosimilar? | If biosimilar, corresponding reference product |
|---|---|---|---|
| Lantus | 100 units/mL in 3-mL cartridges | Reference | — |
| Lantus | 100 units/mL in 10-mL vials | Reference | — |
| Lantus | SoloStar: 100 units/mL in 3-mL prefilled pens | Reference | — |
| Toujeo | 300 units/mL in 1.5-mL prefilled SoloStar pens | Reference | — |
| Abasaglar | 100 units/mL in 3-mL cartridges | Biosimilar | Lantus, 100 units/mL in 3-mL cartridges |
| Abasaglar | KwikPen: 100 units/mL in 3-mL prefilled pens | Biosimilar | Lantus SoloStar, 100 units/mL in 3-mL prefilled pens |
| Product | Presentation | Reference or biosimilar? | If biosimilar, corresponding reference product |
|---|---|---|---|
| Lantus | 100 units/mL in 3-mL cartridges | Reference | — |
| Lantus | 100 units/mL in 10-mL vials | Reference | — |
| Lantus | SoloStar: 100 units/mL in 3-mL prefilled pens | Reference | — |
| Toujeo | 300 units/mL in 1.5-mL prefilled SoloStar pens | Reference | — |
| Abasaglar | 100 units/mL in 3-mL cartridges | Biosimilar | Lantus, 100 units/mL in 3-mL cartridges |
| Abasaglar | KwikPen: 100 units/mL in 3-mL prefilled pens | Biosimilar | Lantus SoloStar, 100 units/mL in 3-mL prefilled pens |
Among other variables, our data sets captured the total number of units (i.e., cartridges, pens, or vials) of each product and each presentation of insulin glargine prescribed by each GP practice as well as the net ingredient cost (i.e., the list price as stated in the Drug Tariff) and actual cost (i.e., actual drug acquisition costs, calculated as the list price, or net ingredient cost, minus an approximate discount plus payment for consumables, containers, and out-of-pocket expenses) for the payer (i.e., the NHS) associated with these. Each GP practice was identified by a unique practice code, and an additional field indicated the CCG to which each GP practice belonged. All prescribing data are also available to the public through a user-friendly interface (OpenPrescribing.net) developed at the University of Oxford (16).
Analyses
We first calculated the total number of units prescribed of each product and presentation, over time, by GPs in England. To understand the market dynamics after the adoption of biosimilars, we calculated the country-level market share of each product and presentation, over time, as the proportion of the total units prescribed. We also analyzed the evolution of the actual cost per unit to understand whether certain changes in the actual cost of treatments have resulted after the commercialization of biosimilars.
Then we estimated the real-world budget impact, or savings, resulting from the adoption of insulin glargine biosimilars by GPs in England. For this, we calculated the total actual costs of insulin glargine in our data sets (i.e., the total costs in the real world) and compared these with the hypothetical scenario in which insulin glargine biosimilars were not available by replacing the cost of biosimilar prescriptions with the cost of the corresponding reference product (Table 1). Additionally, we estimated the missed savings with a threshold analysis (i.e., comparing the total actual costs in the real world with the hypothetical scenario in which insulin glargine biosimilars had a market share of 100%). In other words, to estimate the missed savings, we assumed that access to Abasaglar would have been available to all relevant patients in England after the first GP-prescribed Abasaglar. We consider this to be a valid assumption for this hypothetical scenario analysis because, in theory, all CCGs in England could have made the decision to commission Abasaglar as quickly as the quickest CCG. Because Toujeo is an insulin glargine marketed as 300 units/mL, compared with the 100 units/mL of the other insulin glargine products, it was not considered for potential substitution in the threshold analysis.
We created choropleth maps to observe the uptake of insulin glargine biosimilars at the CCG level and to visually assess whether differences exist across CCGs. The choropleth maps presented here show the proportion of Abasaglar KwikPen from the total 3-mL prefilled pen prescriptions (i.e., the sum of Abasaglar KwikPen and Lantus SoloStar prescriptions). We downloaded CCG boundary data from the Open Geography portal of the Office of National Statistics; boundary data from 2015 were used due to the lack of completeness of the 2016, 2017, and 2018 data sets. Finally, to better understand the reasons that may be driving potential differences in the uptake of insulin glargine biosimilars across CCGs, we submitted two Freedom of Information (FOI) requests to two CCGs with markedly different adoption rates. With these FOI requests, we asked whether the CCGs had put in place any policies to incentivize the use of biosimilars and insulin glargine biosimilars in particular.
Results
Our results show that Lantus SoloStar was the insulin glargine product most frequently prescribed by GPs in England, with nearly 6,000,000 units prescribed in 2018, compared with the nearly 1,400,000 units of Lantus cartridges, the second most prescribed product (Table 2). Nevertheless, the market share of Lantus SoloStar decreased substantially, from 80% just before the introduction of Toujeo and Abasaglar (particularly KwikPen) in late 2015 to 60% in December 2018 (Table 2; see the Supplementary Material for further information). By December 2018, Toujeo and Abasaglar KwikPen reached a market share of ∼10% and 8%, respectively (see the Supplementary Material).
Total units, total costs, and actual cost per unit of insulin glargine in primary care in England since the introduction of insulin glargine biosimilars
| 2015 | 2016 | 2017 | 2018 | Total: 2015–2018 | |
|---|---|---|---|---|---|
| Total units | |||||
| Abasaglar, 100 units/mL cartridges (3 mL) | 214 | 11,973 | 40,406 | 64,086 | 116,679 |
| Abasaglar KwikPen, 100 units/mL pens (3 mL) | 1,131 | 66,540 | 287,331 | 571,320 | 926,322 |
| Lantus, 100 units/mL cartridges (3 mL) | 1,569,162 | 1,616,568 | 1,518,500 | 1,382,872 | 6,087,102 |
| Lantus, 100 units/mL vials (10 mL) | 48,709 | 47,804 | 42,045 | 36,882 | 175,440 |
| Lantus SoloStar, 100 units/mL pens (3 mL) | 6,395,711 | 6,493,615 | 6,257,381 | 5,959,478 | 25,106,185 |
| Toujeo, 300 units/mL pens (1.5 mL) | 9,703 | 183,632 | 485,377 | 818,117 | 1,496,829 |
| Total costs (£) | |||||
| Abasaglar, 100 units/mL cartridges (3 mL) | 1,396 | 78,287 | 264,164 | 419,789 | 763,636 |
| Abasaglar KwikPen, 100 units/mL pens (3 mL) | 7,381 | 435,216 | 1,878,904 | 3,743,194 | 6,064,695 |
| Lantus, 100 units/mL cartridges (3 mL) | 12,046,224 | 12,425,823 | 11,674,629 | 10,014,635 | 46,161,311 |
| Lantus, 100 units/mL vials (10 mL) | 1,382,180 | 1,358,124 | 1,194,763 | 987,773 | 4,922,840 |
| Lantus SoloStar, 100 units/mL pens (3 mL) | 49,102,727 | 49,916,502 | 48,112,182 | 43,147,088 | 190,278,499 |
| Toujeo, 300 units/mL pens (1.5 mL) | 99,124 | 1,878,858 | 4,965,681 | 8,386,202 | 15,329,865 |
| Actual cost per unit (£) | |||||
| Abasaglar, 100 units/mL cartridges (3 mL) | 6.53 | 6.54 | 6.54 | 6.55 | — |
| Abasaglar KwikPen, 100 units/mL pens (3 mL) | 6.53 | 6.54 | 6.54 | 6.55 | — |
| Lantus, 100 units/mL cartridges (3 mL) | 7.68 | 7.69 | 7.69 | 7.24 | — |
| Lantus, 100 units/mL vials (10 mL) | 28.38 | 28.41 | 28.42 | 26.78 | — |
| Lantus SoloStar, 100 units/mL pens (3 mL) | 7.68 | 7.69 | 7.69 | 7.24 | — |
| Toujeo, 300 units/mL pens (1.5 mL) | 10.22 | 10.23 | 10.23 | 10.25 | — |
| 2015 | 2016 | 2017 | 2018 | Total: 2015–2018 | |
|---|---|---|---|---|---|
| Total units | |||||
| Abasaglar, 100 units/mL cartridges (3 mL) | 214 | 11,973 | 40,406 | 64,086 | 116,679 |
| Abasaglar KwikPen, 100 units/mL pens (3 mL) | 1,131 | 66,540 | 287,331 | 571,320 | 926,322 |
| Lantus, 100 units/mL cartridges (3 mL) | 1,569,162 | 1,616,568 | 1,518,500 | 1,382,872 | 6,087,102 |
| Lantus, 100 units/mL vials (10 mL) | 48,709 | 47,804 | 42,045 | 36,882 | 175,440 |
| Lantus SoloStar, 100 units/mL pens (3 mL) | 6,395,711 | 6,493,615 | 6,257,381 | 5,959,478 | 25,106,185 |
| Toujeo, 300 units/mL pens (1.5 mL) | 9,703 | 183,632 | 485,377 | 818,117 | 1,496,829 |
| Total costs (£) | |||||
| Abasaglar, 100 units/mL cartridges (3 mL) | 1,396 | 78,287 | 264,164 | 419,789 | 763,636 |
| Abasaglar KwikPen, 100 units/mL pens (3 mL) | 7,381 | 435,216 | 1,878,904 | 3,743,194 | 6,064,695 |
| Lantus, 100 units/mL cartridges (3 mL) | 12,046,224 | 12,425,823 | 11,674,629 | 10,014,635 | 46,161,311 |
| Lantus, 100 units/mL vials (10 mL) | 1,382,180 | 1,358,124 | 1,194,763 | 987,773 | 4,922,840 |
| Lantus SoloStar, 100 units/mL pens (3 mL) | 49,102,727 | 49,916,502 | 48,112,182 | 43,147,088 | 190,278,499 |
| Toujeo, 300 units/mL pens (1.5 mL) | 99,124 | 1,878,858 | 4,965,681 | 8,386,202 | 15,329,865 |
| Actual cost per unit (£) | |||||
| Abasaglar, 100 units/mL cartridges (3 mL) | 6.53 | 6.54 | 6.54 | 6.55 | — |
| Abasaglar KwikPen, 100 units/mL pens (3 mL) | 6.53 | 6.54 | 6.54 | 6.55 | — |
| Lantus, 100 units/mL cartridges (3 mL) | 7.68 | 7.69 | 7.69 | 7.24 | — |
| Lantus, 100 units/mL vials (10 mL) | 28.38 | 28.41 | 28.42 | 26.78 | — |
| Lantus SoloStar, 100 units/mL pens (3 mL) | 7.68 | 7.69 | 7.69 | 7.24 | — |
| Toujeo, 300 units/mL pens (1.5 mL) | 10.22 | 10.23 | 10.23 | 10.25 | — |
The analysis of actual costs of insulin glargine showed that these have been maintained relatively steady in the last few years, with a considerable reduction in the case of Lantus in 2018 (Table 2; see the Supplementary Material for further information). The actual cost of Lantus SoloStar to the English NHS (£7.01 per pen) was, in December 2018, closer to that of Abasaglar KwikPen (£6.55 per pen) than ever before, which represents savings of £0.46 per pen (or 6.6%) with Abasaglar KwikPen compared with Lantus SoloStar.
Table 3 reports the budget impact, or savings, associated with the introduction of insulin glargine biosimilars in primary care in England. The total savings with Abasaglar between October 2015 (the time of the first captured Abasaglar prescription) and December 2018 was ∼£900,000, most of which was due to the adoption of KwikPen. The total missed savings in England, however, amounted to ∼£25.6 million. With this, the results showed that the percentage of savings realized with the introduction of insulin glargine biosimilars was 3.42% of the total savings that could have been achieved in the period analyzed.
Real-world savings and missed savings with insulin glargine biosimilars in primary care in England
| 2015 | 2016 | 2017 | 2018 | Total: 2015–2018 | |
|---|---|---|---|---|---|
| Savings (£) | |||||
| Savings with Abasaglar cartridges | 246 | 13,744 | 46,488 | 44,316 | 104,794 |
| Savings with Abasaglar KwikPen | 1,303 | 76,278 | 330,346 | 393,208 | 801,135 |
| Total savings with Abasaglar | 1,549 | 90,022 | 376,834 | 437,524 | 905,929 |
| Missed savings (£) | |||||
| Missed savings with Abasaglar cartridges | 450,257 | 1,855,735 | 1,747,052 | 956,270 | 5,009,314 |
| Missed savings with Abasaglar KwikPen | 1,837,423 | 7,443,899 | 7,194,153 | 4,101,579 | 20,577,054 |
| Total missed savings with Abasaglar | 2,287,680 | 9,299,635 | 8,941,205 | 5,057,849 | 25,586,369 |
| Percentage of savings realized (%) | 0.07 | 0.96 | 4.04 | 7.96 | 3.42 |
| 2015 | 2016 | 2017 | 2018 | Total: 2015–2018 | |
|---|---|---|---|---|---|
| Savings (£) | |||||
| Savings with Abasaglar cartridges | 246 | 13,744 | 46,488 | 44,316 | 104,794 |
| Savings with Abasaglar KwikPen | 1,303 | 76,278 | 330,346 | 393,208 | 801,135 |
| Total savings with Abasaglar | 1,549 | 90,022 | 376,834 | 437,524 | 905,929 |
| Missed savings (£) | |||||
| Missed savings with Abasaglar cartridges | 450,257 | 1,855,735 | 1,747,052 | 956,270 | 5,009,314 |
| Missed savings with Abasaglar KwikPen | 1,837,423 | 7,443,899 | 7,194,153 | 4,101,579 | 20,577,054 |
| Total missed savings with Abasaglar | 2,287,680 | 9,299,635 | 8,941,205 | 5,057,849 | 25,586,369 |
| Percentage of savings realized (%) | 0.07 | 0.96 | 4.04 | 7.96 | 3.42 |
Our analyses demonstrated that the uptake of insulin glargine biosimilars varied substantially across CCGs, but no clear geographical pattern emerged (Fig. 1). The NHS Swindon CCG was the quickest CCG to adopt insulin glargine biosimilars, with an average of 7.4% of use of Abasaglar KwikPen (from the total 3-mL prefilled pen prescriptions) across GP practices in December 2015. The uptake of Abasaglar KwikPen, over time, was unequal across England, with NHS North and West Reading CCG leading the uptake of Abasaglar KwikPen in December 2016 (26.9%) and NHS Kernow CCG in December 2017 (40.2%). In December 2018, NHS Swindon CCG and NHS Kernow CCG were the CCGs with the highest use of insulin glargine biosimilars (53.3% and 51.4%, respectively), followed by NHS Berkshire West, which included the former NHS North and West Reading CCG (41.2%) (Fig. 1; the “missing data” were not actually missing, but due to the evolution of CCGs and lack of complete boundary data associated with these changes, we were not able to present all the data in the map). On the other hand, the use of Abasaglar KwikPen in December 2018 was 0% in 9 CCGs, with a total of 28 CCGs below a 1% use. Choropleth maps of Abasaglar KwikPen uptake in December 2015, December 2016, and December 2017 are available in the Supplementary Material.
Market share of Abasaglar KwikPen, from the total 3-mL prefilled pen prescriptions, in December 2018.
Market share of Abasaglar KwikPen, from the total 3-mL prefilled pen prescriptions, in December 2018.
One FOI request was submitted to NHS Kernow CCG (one of the CCGs with the highest insulin glargine biosimilar use), and the response to our request stated that NHS Kernow CCG has neither incentivized an increase in the use of biosimilars in Cornwall nor, more specifically, has it incentivized the uptake of biosimilar insulin glargine. It was confirmed, however, that the CCG supported the review of all people with diabetes who are prescribed analog insulin and a switch to a biosimilar if appropriate and acceptable to the individual. We submitted another FOI request to South Sefton CCG (one of the CCGs with an uptake rate of insulin glargine biosimilars of 0%), but the response only mentioned that agreeing on policies around biosimilars is part of the CCG’s work plan for 2019–2020.
Conclusions
Summary of Findings
Our analyses showed that the total savings with Abasaglar between October 2015 (the time of the first Abasaglar prescription) and December 2018 was ∼£900,000, which represents 3.42% of the total savings that could have been realized in the period analyzed. In other words, savings of ∼£25.6 million on drug acquisition costs were missed from not using insulin glargine biosimilars in that period, assuming full access to Abasaglar in primary care in England, ceteris paribus. The extent of future savings and missed savings will depend on a number of variables, such as price adjustments, the entry of competitor biosimilars and the subsequent market dynamics, and incentives to clinicians and patients, but our results provide a valuable idea of the range of future realized and missed savings. Our analyses also demonstrated that a substantial level of variation in the use of insulin glargine biosimilars exists across CCGs in England, ranging from 0 to 53.3% in the case of Abasaglar KwikPen. The information gathered from the two FOI requests did not provide sufficient clarification about why this level of variation exists between one of the CCGs with the highest insulin glargine biosimilar use and one of the CCGs with the lowest use.
Strengths and Limitations
In recent years, the U.K. Government has made public a large amount of data to encourage analysis and innovation, including detailed, monthly data sets on treatments prescribed by GP practices in England (14). Using these rich data sets represents the key strength of our study, and with this, we not only shed light on an issue with relevant implications at the policy level but also contributed toward the very purpose of making all of these data sets available to the public. In addition, our study provides a framework that can be updated regularly to evaluate the uptake and variation of any particular treatment across GPs in England and to promote policies that ensure equitable access to health care. This framework could be used at the national, regional, or local level.
Our study, however, also had certain limitations. A number of medical associations and regulatory agencies, including the Medicines and Healthcare Products Regulatory Agency in the U.K., have stated that all biological medicines must be prescribed by brand name to avoid automatic substitution at the pharmacy level (2). The GP practice–level prescribing data, however, captured a considerable number of insulin glargine prescriptions (13.3%) without a brand name associated with the given prescription record. For this reason, and to prevent potential errors, we focused on those records with an associated brand name and removed the rest from our analyses. Therefore, our results are likely to be an underestimation of the real budget impact as well as an underestimation of the total missed savings. On the other hand, our study only considered drug acquisition costs and did not account for potential costs of switching patients from originators to biosimilars, such as the costs of regularly monitoring blood glucose levels or potential nocebo effects, meaning that the savings we estimated will likely be reduced in practice by these extra costs (17,18). Switching costs have been shown to be substantial in other disease areas, but managed biosimilar switch programs have also been successfully implemented in the U.K. despite this additional burden (19–21). Although switching costs are not expected to exceed the savings realized by an increased use of insulin glargine biosimilars, the lack of data on switching prevents us from assessing this in detail.
The lack of patient-level data was another limitation, because we were not able to assess any potential associations between the use of specific treatments (e.g., the uptake of insulin glargine biosimilars) and patients’ characteristics. This was an unavoidable limitation resulting from the nature of the data set.
Finally, the responses to our FOI requests did not provide the extent of detail we were hoping for, thus limiting our ability to understand whether policies for access to insulin glargine biosimilars varied across CCGs and what may be driving the observed differences.
Policy-Level Implications
As part of the U.K. Government’s Long Term Plan (LTP) for the NHS, the NHS announced savings of £700 million were achieved in 2018–2019 as a consequence of maximizing the use of best value generic and biological treatments (22). A significant proportion of this figure came from the uptake of adalimumab biosimilars after Humira lost its patent protection in October 2018: the increased use of adalimumab biosimilars delivered savings of ∼£110 million in 2018–2019 (22). Other biologic treatments also contributed substantially to these savings: £45 million with the best value rituximab, £36 million with the best value etanercept, £32 million with the best value infliximab, and £24 million with the best value trastuzumab (22). Unfortunately, no data were reported on savings with insulin glargine biosimilars in primary or secondary care, which was in line with the results from the CCG Biosimilar National Questionnaire suggesting that most CCGs only develop plans for the most frequently used biosimilars in secondary care. The reality is that savings with insulin glargine biosimilars in primary care (∼£900,000 between October 2015 and December 2018) are minor compared with those of the best value adalimumab, rituximab, etanercept, infliximab, and trastuzumab. It is also true, however, that the savings missed from not using insulin glargine in primary care in 2018 (£5.1 million) (Table 3) are at the level of the savings realized in 2018–2019 with the uptake of generics such as caspofungin (£8.5 million), valganciclovir (£3.8 million) or voriconazole (£7.0 million), according to the NHS LTP report (22).
Regulatory and reimbursement agencies affirm that no clinically meaningful differences exist between originators and biosimilars, and at the same time, the NHS has encouraged the use of the best value biologic treatment (1,23,24). In spite of this, considerable barriers exist for the adoption of biosimilars and, as a consequence, for achieving all potential savings (13,25). For instance, a survey in U.K. hospitals concluded that physicians’ perception of efficacy and safety was crucial, showing that only 30% of diabetologists had no concern for starting treatment with insulin glargine biosimilars or switching treatment for patients already on treatment (25). This could be justified by the incidents reported when switching patients from Lantus to Abasaglar, posing an additional hurdle for a smooth uptake of insulin glargine biosimilars and the concerns regarding medical liability (26–29). In addition, the ABCD stated there is a real problem in diabetes care regarding the “lack of knowledge” among health care professionals about the characteristics of existing treatments, with biosimilar insulins adding to this complexity (5). The sparse evidence base comparing biosimilar insulins with their originators is a significant barrier contributing to this lack of knowledge (28,29).
Patients’ perceptions and resistance to changing treatments may also play a role. A patient survey conducted in Canada demonstrated that patients remain resistant to switching to biosimilars, which has implications on treatment adherence and subsequently on patient outcomes (30). Despite differences between the Canadian and English health care systems, similar results may be expected in England. With all this, better and proactive education programs about biosimilars for both clinicians and patients have been considered crucial to increase the use of biosimilars (28).
Other barriers may include the costs of switching patients from Lantus to an insulin glargine biosimilar (e.g., having prescribing pharmacists in GP practices, referring the patient to a specialist in a hospital, or having to closely monitor patients’ blood glucose) and the fact that CCGs may simply focus on those molecules with the greatest potential to deliver substantial savings in a short period of time (13). The launch of Toujeo and its quick uptake may have also played a role in limiting the use of insulin glargine biosimilars.
According to the Commissioning framework for biological medicines, there is a significant opportunity to further benefit from biosimilar medicines if action is taken across the country and best practice is implemented (2). For this reason, and to optimize the use of NHS resources, implement best practice, and reduce unwarranted variation, Regional Medicines Optimisation Committees (RMOCs) were established in 2016 (23). The role of RMOCs is to provide advice to commissioners (such as CCGs) and providers, and in turn, these are expected to follow and implement the RMOC’s advice (23). One example of advice is the one released by the Guilford and Waverley RMOC in December 2016, which recommended the use of “Abasaglar® for new patients who would previously have been initiated on Lantus®” (31). Initiating diabetes patients on insulin glargine biosimilars seems to be an accepted approach that budget holders across England are promoting; however, there is more resistance to switch stable patients currently on Lantus (13,27). A survey in U.K. hospitals revealed that diabetologists in secondary care believe that switching to insulin glargine biosimilars should be done by GPs in primary care, because it would be the primary care budget that would benefit (13). But switching patients and realizing savings will be limited if the appropriate structure and incentives do not exist. Apart from following the advice given by RMOCs, the NHS and CCGs may need to put in place certain incentive schemes to promote the uptake of biosimilars in primary care, such as reinvesting the savings within the CCG or practice that realized the savings, putting in place supporting staff to incentivize the uptake of insulin glargine biosimilars, as was done by the Mid Essex CCG, or extending the Quality and Outcomes framework to include items on the use biosimilars in primary care (2,13). In case a wider implementation of incentive schemes such as these proves successful, current manufacturers would have to adapt their strategies to remain key players in the field. Further pricing adjustments would be likely, and biosimilar manufacturers would need to adapt their production levels to avoid potential drug shortages and maintain their reputation.
The atlas developed by the NHS Business Service Authority (i.e., Medicines Optimisation Dashboard) illustrates that variations in prescribing rates exist across NHS Trusts in the use of the biosimilars of etanercept, infliximab, rituximab, and trastuzumab (32). For instance, the uptake of infliximab biosimilars in July 2018 ranged from 51% (in Bradford Teaching Hospitals NHS Foundation Trust) to 100% in several trusts; also in July 2018, the uptake of etanercept, rituximab, and trastuzumab varied between 0 and 100% across NHS trusts. Similar results were reported by other studies that used the GP practice-level prescribing data to analyze variations (across CCGs) in other products’ prescription rates, indicating that inequalities in prescription rates may be widespread across primary care in England (33–35). Our results corroborate this reality: it seems that patients in certain areas are more likely to be initiated on or switched to insulin glargine biosimilars compared with other areas. This adds further evidence to the postcode lottery issue in England (36,37). Although we tried to decipher some of the key determinants of the variation in the uptake of insulin glargine biosimilars by submitting two FOI requests, this was not the key focus of our analysis and remains a topic of future research.
Summary
Two key conclusions can be drawn from this analysis: firstly, there are considerable savings that are not being realized, and secondly, there is substantial variation in the use of insulin glargine biosimilars across CCGs in England. We believe that the results from our analyses will encourage decision makers in England, as well as health care managers in other settings, to promote a coordinated approach to the use of best value treatments in primary care and to reevaluate variation in treatment patterns across GP practices and CCGs—because switching to biosimilars is not just about generating savings but also about creating budget headroom for an increased access to best value treatments for all patients.
This article contains supplementary material online at https://doi.org/10.2337/figshare.12240668.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. I.A., E.S.-I., K.M., and J.M.C. designed the work. I.A. collected, cleaned, and analyzed the data. I.A. drafted the manuscript. All authors revised and approved the work before submission. I.A. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
