Pregabalin: what the patent litigation means for doctors and drug companiesBMJ 2018; 361 doi: https://doi.org/10.1136/bmj.k2318 (Published 07 June 2018) Cite this as: BMJ 2018;361:k2318
- Correspondence to: B Goldacre
Generic prescribing is near universal in the UK, taught in all medical schools, and recommended (with few exceptions) in the BNF.1 It has substantial clinical and cost benefits: generic drugs are usually cheaper than branded products and the generic names make it easier for clinicians and patients to know which class each drug belongs to, providing a harmonised nomenclature that can help reduce the risk of drug errors.
However, in an unprecedented step, in 2015, NHS England wrote to all clinical commissioning groups (CCGs) instructing them to ensure that doctors prescribe pregabalin as branded Lyrica when it was being used to treat pain. The letter, which caused clinicians much concern, was the most visible output from a complex and extensive web of interlocking legal battles about patent protection (box 1). Collectively, these cases have enormous financial consequences for the NHS and raise important issues around medical autonomy, incentives for innovation, the role of regulators, and the entire legal framework around new uses for old drugs.
How the pregabalin litigation unfolded
17 May 2013: Pfizer product patent on pregabalin expires (EP0641330)
14 October 2013: Lapse of supplementary protection certificate (SPC) because of non-payment of fees noted on Register of Patents in UK
19 March 2014: Generic firm Actavis becomes aware that SPC has not been activated
24 June 2014: Generic firm Mylan begins patent revocation action against Pfizer’s second medical use patent (EP0934061)
8 July 2014: Data exclusivity on Lyrica pregabalin marketing authorisation expires, meaning a generic firm can use Pfizer trial data to obtain a marketing authorisation
9 July 2014: Actavis applies for marketing authorisation for their form of pregabalin, Lecaent
8 December 2014: Pfizer starts patent infringement claim against Actavis, Actavis counter-claims for invalidity
16 February 2015: Actavis’ marketing authorisation for Lecaent granted
17 February 2015: Actavis launches Lecaent pregabalin, with “skinny label”
26 February 2015: Patents Court orders NHS England to issue prescribing instructions
10 September 2015: Patents Court finds Pfizer second medical use patent is invalid for most of its claims, including neuropathic pain
13 October 2016: Court of Appeal upholds finding of invalidity for neuropathic pain claims of second medical use patent
16 July 2017: Pfizer second medical use patent for neuropathic pain expires
17 May 2018: Pfizer SPC would have expired if Pfizer had chosen to pay fees.
All information is from the judgments, which can be read here. http://www.bailii.org/ew/cases/EWHC/Patents/2015/2548.html (main judgment), http://www.bailii.org/ew/cases/EWHC/Patents/2015/485.html (judgment ordering prescribing instructions), http://www.bailii.org/ew/cases/EWCA/Civ/2016/1006.html (Court of Appeal judgment)
Authorisation of pregabalin
Pregabalin is sold by Pfizer under the brand Lyrica. It was initially authorised in the European Union for treatment of epilepsy and neuropathic pain but licensing was later extended to generalised anxiety disorder. The product patent protection for pregabalin, based on its activity as an anticonvulsant, expired on 17 May 2013. However, Pfizer later sought a “second medical use” patent for pregabalin in pain, which expired on 16 July 2017. (The patent is actually in the name of Warner Lambert, a Pfizer subsidiary, but for convenience no distinction will be made here).
Mystery of supplementary protection certificates
This second patent should have been a minor footnote in legal history and not affect prescribing habits or cost. However, unexplained behaviour by Pfizer created a complicated legal situation for generic manufacturers and prescribers and resulted in the company forfeiting billions of pounds in revenue.
In 2004, Pfizer applied for, and was granted, supplementary protection certificates2 based on the earlier product patent for pregabalin across Europe; this is common practice. These certificates would have lasted for five years and therefore extended Pfizer’s market exclusivity to May 2018 for all indications. To activate the certificates Pfizer had to pay a modest administrative fee (a few thousand pounds in the UK) when the underlying patent expired in 2013.3
Inexplicably, Pfizer never paid these fees, in any of the territories in which certificates were granted, so they were never activated. A company with patent lawyers across Europe informing it that the payment was due is unlikely to have allowed the certificates to lapse inadvertently. Pfizer has not explained this highly unusual decision, and the speculative reasons suggested by various commentators are not convincing; it remains a mystery.4 In March 2014 the generic manufacturer Actavis became aware through Patent Office records that Pfizer had not paid the administrative fees. Other manufacturers were therefore able to market generic pregabalin, at least for indications other than pain, after July 2014.5
Litigation and “skinny labels”
Actavis was one of the first companies to apply for and be granted a generic marketing authorisation on pregabalin, with Lecaent. Pfizer promptly sued for patent infringement and demanded that Actavis ensure that Lecaent was not used to treat pain. Lecaent was not licensed for use in pain; instead, it had a “skinny label” covering only those indications for which Pfizer no longer held a patent (epilepsy and generalised anxiety disorder).
A skinny label is a new and still relatively rare phenomenon; the general rule is that a generic form of a drug must be labelled the same way as the reference drug. The label presented a new problem around policing. Manufacturers do not have a mechanism to control how doctors use their drugs, or what medical problems they use them for. Similarly, they cannot control pharmacists’ dispensing behaviour. In any case, pharmacists are unlikely to know the indication for a prescription: although some GP computer systems have the ability to include the indication, the feature is not commonly used, and doctors may not have a single indication in mind when they prescribe.
The judge at a hearing in February 2015 thought he had found a solution to this problem. When a drug is prescribed by brand, the pharmacist must dispense that precise brand; when a drug is prescribed by generic name, the pharmacist may dispense any brand. At Pfizer’s urging—but without much argument from Actavis, NHS England, or the Department of Health—the Patents Court ordered NHS England to instruct CCGs that they must tell doctors to use the brand name Lyrica when pregabalin was being prescribed for neuropathic pain, although the generic name should be used for other indications.67 CCGs were also required to amend any prescribing decision support software (such as Scriptswitch, or OptimiseRx) to show this message. The NHS Business Services Authority was additionally ordered to instruct all community pharmacies that they should ensure, as far as possible, that Lyrica is dispensed if the prescription is known to be for pain.
This was an unprecedented situation in many ways. Because the possibility of launching a skinny label product had become apparent only when Pfizer’s extension certificates were unexpectedly not activated, there was no time for generic companies to clear the path for generic launch—for example, by challenging the validity of Pfizer’s second medical use patent for pain. Furthermore, the indication remaining under patent (pain) was believed to represent a much larger market than the other indications no longer protected by patent, increasing the financial importance of policing the patent’s boundary.
It was not just manufacturers who faced legal action. In correspondence in late 2014 Pfizer suggested that doctors, pharmacists, and CCGs could be individually liable for patent infringement, or procuring patent infringement, if pregabalin was prescribed generically for neuropathic pain. This potentially exposed large numbers of NHS clinicians to expensive and time consuming litigation. The Patents Court subsequently granted a declaration in September 2015 that Actavis, wholesalers of Lecaent, doctors, pharmacists, and patients did not infringe the patent and declared that several communications from Pfizer to pharmacists, CCGs, and the BMA constituted unjustified threats under these particular circumstances.5 However, the situation remains unclear, and the Supreme Court decision on Pfizer’s neuropathic pain patent, due soon, will hopefully clarify who may be held liable for infringement of a second medical use patent, as well as under what circumstances the generic manufacturer might infringe.
How GPs responded, and the cost to the NHS
We were keen to establish how general practitioners responded to letters from NHS England and threats from Pfizer. In our accompanying paper in BMJ Open we present data showing that overall GPs did not comply with NHS England’s instructions.8 About 70% of pregabalin prescriptions are estimated to be for pain, but during the period of litigation only 11% of practices prescribed Lyrica at this level, and no CCGs did so. Although most practices did change their prescribing behaviour, they used Lyrica infrequently.
The generic prescribing did not result in cost savings for the NHS because the reimbursement price for generic pregabalin during this period was the same as branded Lyrica. The reasons for this lie in one final interesting twist. The processes used by the Department of Health to classify and price medicines in the Drug Tariff are not made public, but it is likely that it did not wish to recategorise pregabalin as generic while the pain patent was still being asserted by Pfizer. However, Pfizer’s pain patent was itself contested, as we now discuss.
Did Pfizer own pregabalin for pain?
In September 2015 the Patents Court held that Pfizer’s secondary use patent for pregabalin in pain was partially invalid, on the grounds that the data provided did not show that pregabalin was effective and supported its use only in specific subtypes of pain (which were, notably, outside the marketing authorisation).59 This had important implications for the arguments over pregabalin and skinny labels, and for price. If the pain indication was no longer protected, pregabalin would effectively be entirely off-patent, generics could be marketed freely, and any generic could be prescribed for any use. This finding was upheld by the Court of Appeal in October 2016, and a further appeal to the Supreme Court was heard in February 2018, although the judgment is still awaited.
This ruling will be important because the court order in 2015—requiring NHS England to ask doctors to prescribe Lyrica by brand for pain—came with conditions. Specifically, there was a “cross-undertaking in damages,” whereby Pfizer agreed to compensate NHS England for any financial loss should the court order turn out to be unjustified—for example, if the patent for pain was found ultimately to be invalid or not infringed.
The price of pregabalin dropped in August 2017, when the second medical use patent expired. If Pfizer had not appealed the loss of the second medical use patent for pain in 2015, then pregabalin would instead have become generic at that time and the price drop observed in August 2017 would have occurred in October 2015. We estimate that the delay in the price drop resulted in total excess costs to the NHS of £502m,8 representing a substantial proportion of the entire NHS primary care prescribing spend (roughly £9bn in England for 2017).10 If the Supreme Court upholds the decision that Pfizer’s pain patent is invalid, NHS England should be able to recoup some excess prescribing costs from Pfizer.
Skinny labels and generic prescribing: can exceptions be justified?
This complex set of interlocking disputes has led to an extraordinary situation in which doctors have been instructed by the NHS to depart from their usual practice of generic prescribing. Furthermore, this has been done for a legal reason—to prevent patent infringement—rather than for a clinical reason. As authors (a doctor, a pharmacist, and a patent attorney) we have had extensive and repeated discussion, disagreement, and partial changes of heart about whether this is justifiable.
On the one hand, breaching the principle of generic prescribing can be viewed as an unwarranted and inappropriate intrusion into the clinical independence of doctors. It also has wider cost and safety benefits throughout medicine. Given that most GPs seem to have independently decided not to follow NHS England’s instruction, the medical community may agree with this position.
On the other hand, innovation must be incentivised, and new uses for old drugs often represent innovation. If a drug company cannot expect to be able to enforce a second medical use patent, it has no motivation to develop and seek marketing authorisation for follow-on indications. If skinny labels cannot be policed, or if doctors do not cooperate in this policing, then generic availability may be delayed until all such additional patents have expired. The Court of Appeal has made it clear that generic companies commit patent infringement when it is “foreseeable” that their product will be used for the patented indication. If the patented indication cannot be separated from uses where the patent has expired, the financial liability for the generic company would greatly exceed any foreseeable profit and they would not try to market a lower cost generic. This is the very situation that the skinny label system was designed to rectify.
Notably, the view of regulators here is also unclear. The General Medical Council has issued guidance that unlicensed drugs should be prescribed only when there is a “special need” that “could not be taken to encompass” situations when there was “a licensed alternative.”11 Generic pregabalin was not licensed for pain: its use in pain is therefore technically off-label and against the GMC guidance.12 However, with a focus on efficacy and safety, many doctors would distinguish between genuine off-label prescribing, where the medicine is not licensed for a particular indication at all, and the situation presented by pregabalin in pain, where the medicine has been licensed for the indication but the particular brand used has not. This has been termed cross-label prescribing.13 Moreover, the GMC has recently confirmed that widespread off-label prescribing can sometimes be justified, as in the case of bevacizumab, which is included in NICE guidance for the treatment of macular degeneration despite the manufacturer not seeking a marketing authorisation for that condition.1415
As final additional context, the peculiarities of NHS drug pricing have recently led to a resurgence of branded prescribing: specifically, many CCGs now encourage cost savings through prescription of some “branded generics” (specific brands of generic drugs that may be cheaper than the NHS Drug Tariff price for the generic).
The ongoing disputes around pregabalin have exposed tensions between clinical freedom, regulatory oversight, and intellectual property. They have also raised several complex new questions that remain unresolved. Is it practical to collect data on the intent of each prescription? Which organisations, institutions, or professions could use this information to control off-patent use for a skinny label? And what are the consequences for medical innovation, or drug costs, if skinny labels prove to be unviable? We encourage the wider policy and medical community to engage robustly with these issues. With a wider push throughout industry and academia to identify and exploit new uses of old medicines, this situation is likely to recur.
Product patent—A patent with claims to a substance (for example, an active ingredient), not limited to any particular use or application
Second medical use patent—A patent with claims directed to the use of an active ingredient for a particular therapeutic indication; the patent may have multiple claims defining the indication at different levels of generality—eg, “pain” in one claim, “neuropathic pain” in another claim, and “inflammatory pain” in another
Supplementary protection certificate (SPC)—An intellectual property right granted to a patent holder to compensate for the period when a patent for a medicine cannot be exploited because of the time taken to obtain regulatory approval for sale of the medicine. An SPC has a maximum term of five years after expiry of the underlying patent, and the scope of protection is equivalent to that of the underlying patent but restricted to the authorised active ingredient. If the underlying patent is a product patent, then the SPC protects the authorised active ingredient for all indications that are authorised during the lifetime of the SPC
Marketing authorisation—The license granted to a company (in the UK either by the European Medicines Agency or the Medicines and Healthcare Products Regulatory Agency) to allow the company to sell a medicine. The marketing authorisation covers a “summary of product characteristics,” also referred to as a “label,” which includes the medical indications for which the medicine is authorised
Generic marketing authorisation—Marketing authorisation granted under an abbreviated procedure in which it is not necessary to show safety and efficacy of a medicine, provided that it is shown to be equivalent to an existing medicine (the “reference medicine”) that already has a marketing authorisation
Data exclusivity—The period (usually 10 years from the date of the marketing authorisation of the reference drug) during which it is not permitted to obtain a generic marketing authorisation. This is to give the originator company a period of exclusivity in the market
Skinny label—A summary of product characteristics in a generic marketing authorisation that omits one or more indications included in the reference marketing authorisation because of the existence of a patent. The skinny label system was introduced in Europe in 2004 and was designed for the eventuality of patent expiry for only a subset of a medicine’s indications
The patent for pregabalin expired in 2014 for all uses other than for pain
The arrival of a generic version led to a complex set of interlocking and ongoing legal disputes
In 2015 the Patents Court ordered that GPs must specify branded Lyrica when prescribed for neuropathic pain, but doctors largely ignored this instruction
The situation raises questions about how to balance clinical freedom and NHS costs against incentives for drug innovation
Overall, new mechanisms designed to accelerate access to generic drugs could paradoxically end up causing delay
We thank Hayley Gould and Janice Headon for advice on GP and community pharmacy systems; and the OpenPrescribing team at the EBM DataLab in the University of Oxford for obtaining and processing all NHS England primary care prescribing data.
Competing interests: We have read and understood BMJ policy on declaration of interests and declare the following: BG has received research funding from the Laura and John Arnold Foundation, the Wellcome Trust, the Oxford Biomedical Research Centre, the NHS National Institute for Health Research School of Primary Care Research, the Health Foundation, and the World Health Organization; he also receives personal income from speaking and writing for lay audiences on the misuse of science. RC is employed on BG’s grants for the OpenPrescribing project. RC is employed by a CCG to optimise prescribing and has been a paid member of advisory boards for Martindale Pharma, Menarini Farmaceutica Internazionale, and Stirling Anglian Pharmaceuticals. DS is a patent attorney in private practice and a partner in the intellectual property law firm EIP. He and other patent attorneys and solicitors at EIP act for and advise drug companies as patent holders as well as advising and representing in relation to patents held by other entities. DS has written extensively concerning pregabalin in his capacity as member of the IPKat intellectual property law blog and as member of the editorial committee of the Journal of Intellectual Property Law and Practice; both these roles are part of his general professional activity and are not remunerated. All authors conceived the project, drafted the manuscript, contributed to and approved the final manuscript.
Provenance and peer review: Not commissioned; externally peer reviewed.