This year, there have been several more battles in the opioid over-use war in the United States (US). Firstly, Walmart (a large chain of pharmacy retailers in the US) agreed to pay US$ 3.1 billions to a collection of city governments, native American tribes and US states. This follows similarly large settlements reached by Walgreens Boots Alliance and CVS Health [1]. As discussed in last year’s review [2], deep-pocketed retail pharmacy companies are now firmly in the plaintiffs’ range-finders (in addition to their other, long-standing targets, namely, pharmaceutical companies and prescribers). All of these pharmacy chains have denied wrongdoing.

Later in the year, a further step came in the huge litigation concerning oxycodone abuse that was mounted by various civil authorities in the USA against Purdue Pharma; readers may remember that, in 2018, a proposed settlement was rejected by the courts. This litigation has now achieved a court-approved bankruptcy of the company. Furthermore, the owners of the (private) company have now agreed to pay US$6 billions in order to terminate all further civil actions against themselves [3]. A new company (called ‘Knoa Pharma’ [4]) will be formed to salvage the remainder of the old one, and it will be overseen by ‘a public board’. The headlines have been somewhat contradictory, but, whatever the justice of the settlement, one wonders what proportion of these enormous damages will now reach the intended beneficiaries.

In January, a further episode in the US opioid abuse war took place when the (federal) Justice Department filed a lawsuit against a large wholesale drug distributor, i.e., upstream from the retail pharmacy chains mentioned above [5]. The wholesaler has allegedly violated the Controlled Substances Act (1970) by failing to report certain ‘suspicious’ purchases to the Drug Enforcement Administration (DEA). ‘Suspicious’ purchases are defined as sales orders of unusually large size, increased frequency, or from questionably legitimate retailers. This, the Justice Department contends, has been going on for a decade. One retail purchaser (in New Jersey) is alleged to have employed staff with prior, drug-related, criminal convictions, for which the Government is holding the wholesaler responsible. A further complaint is that some suspicious sales of opioids took place in a car park in Florida; one wonders how or why a reputable wholesaler might be the cause of that. Meanwhile, the wholesaler claims that it had, indeed, made the required reports to the DEA, and that it had checked the state licences of all the retailers that it supplied. The defence further claims that only five out of several thousands of their retail clients have been specified in the complaint, with four of whom the wholesaler had ceased doing business before the DEA became involved. The wholesaler has also questioned why the DEA took no prior action, during the alleged decade. Once again, the purist’s legal issue of proximate cause is probably going to be sub-ordinated to the question of what can be extracted from the deepest pockets: the wholesaler in question has annual US$ revenues in 12 figures.

However, there was a glimmer of hope for rationality in the US litigation arena this year. Consolidated lawsuits alleging that the venerable ranitidine was oncogenic were dismissed by a federal judge in Florida [6]. The judge found a lack of “objective, science-based standards for even-handed evaluation of [plaintiffs’] data”, and that “There is no scientist outside this litigation who concluded ranitidine [marketed as Zantac®] causes cancer” [7]. The ruling can be appealed, and a few other cases are still pending in various US state courts. Nonetheless, it is refreshing that, for once, claimants’ anecdotes, innuendos, and dubious statistical gymnastics have not been accepted prima facie in at least one US Federal Court.

The year has seen a boost for those who, in spite of all the clinical trial failures, have clung on to the idea that amyloid might be a druggable target in the treatment of Alzheimer’s Disease (AD). Mid-May saw the announcement by Eli Lilly of their phase III data with donanemab. Initially, the imaging and cognitive scale data were published, followed shortly by activity of daily living findings [8]. This repeat-dose intravenous monoclonal antibody binds to amyloid proteins (possibly including tau), and it is thought that the ligand-antigen complex then stimulates intrinsic immune attack. Regardless of whether that is too simplistic a description, the imaging showed amyloid plaque clearance, the psychological scales (e.g., ADAS-Cog13 and others) showed halting of cognitive decline, and there was corresponding benefit in activities of daily living. At least one effusive newspaper reported this as a ‘wonder drug’ [9]. The study seems to have been properly informed by the well-recognized issue: AD needs to be treated before the horse has bolted. The clinical trials used PET scanning at the screening stage, and the randomisation was stratified for estimated tau concentrations. These data strongly support the use of PET scanning as a surrogate end-point in clinical trials for both patient selection and demonstration of drug efficacy. However, in ordinary clinical practice, PET scan availability would probably be a rate-limiting resource outside the USA (and, if that requirement becomes part of US labelling, then possibly also within it).

The beleaguered, amyloid-is-druggable group were further vindicated when a second humanised monoclonal antibody called lecanemab was compared with a placebo in patients with early AD characterised by mild cognitive impairment and/or mild dementia [10]. The phase III clinical trial had a sound and simple design. This time there was no PET scan stratification, and a straightforward, parallel-group, 1:1 randomisation was used (n = 1795 patients). The recruited treatment groups were well-balanced for ApoE4 demography as well as concomitant treatments. After 18 months therapy (IV, every 2 weeks), the active treatment was superior to the placebo, and this was consistent across all subsets of the Clinical Dementia Rating-Sum of Boxes (a standard clinical psychology tool). A sub-study using PET (n = 698 patients) demonstrated reduction in brain amyloid burden, albeit with corresponding signs of oedema and effusions (mostly very small) in twice as many active as placebo-treated patients (26.4% vs 12.6%). An excess of adverse events termed ‘infusion reactions’ was attributable to the active therapy; these were both preventable and reversible with NSAIDs, antihistamines and/or glucocorticoids. It must be remembered that scores from clinical psychology tools are also essentially surrogate endpoints for ‘hard’ clinical outcomes such as survival and development of advanced AD, and that 18 months’ follow-up is probably too short to measure the latter. A long-term, open-label study continues.

However, the beleaguered group did not get it all their way this year. Two studies of crenezumab, with a similar proposed mechanism of action, and similar phase 3 trial designs, were halted after a planned interim analysis due to futility [11].

The US Congress can appoint investigational committees into the actions of any government department; that includes the Food and Drug Administration (FDA). One such committee has been looking into how FDA handled the Biological Licence Application (BLA) for aducanumab (another intravenous monoclonal antibody) [12]. Again, aducanumab targets aggregated amyloid beta, but it already has marketing approval for treating patients with mild to moderate dementia [13]. The product also has unresolved neurological safety signals (cf. lecanemab, see above). The congressional investigation found that the FDA approval process was “rife with irregularities”. These irregularities included dissent from its own statisticians about the pivotal efficacy data, and that the Agency’s advisory committee (comprising external experts) had recommended that the BLA should be rejected (or, in regulatory parlance: ‘referred back to the Sponsor with a complete response’). Meanwhile, during the hearings, and as might be expected, the US Congressmen did not miss the opportunity to criticise the manufacturer for the price of the product (recently reduced to $28,800 per annum), even though drug prices are not actually regulated by FDA. Large medical centres and Health Maintenance Organisations in the US often have their own in-house pharmacoeconomists, an equivalent of the National Institute of Health and Care Excellence (NICE) in the United Kingdom (UK), and generally decide for themselves whether products can or cannot be prescribed.

This year, both the High Court and the Supreme Court in the UK have been troubled by cases involving informed consent. Briefly, old law revolved around a case known as Bolam [14] which created a legal test (the ‘professional practice test’) for medical negligence: if a prescription or a surgical procedure that caused harm could reasonably have also been recommended by a responsible body of medical opinion, even if there might have been others with a contrary opinion, then (usually) it is not considered negligence. Bolam was superseded by Montgomery [15] for informed consent where the Supreme Court decided that:

“…the professional practice test did not apply to a doctor’s advisory role in discussing with the patient any recommended treatment and possible alternatives, and the risks of injury which may be involved”.

In July, the High Court heard Powell, which returned to this subject [16]. The patient’s clinical course in this case was fairly straightforward, while the legalities were more complicated. The clinical scenario was that a patient with multiple co-morbidities (making her a somewhat poor surgical candidate) had an infected knee implant. Her long-standing surgeon (four previous operations) recommended debridement, antibiotics, irrigation and replacement (DAIR) as a single procedure, with the caveat that, depending upon what was found intra-operatively, he might instead choose to close the surgical wound without the replacement, wait for the infection to clear, and, at some later date, offer a second operation for the replacement arthroplasty. In the event, DAIR was done. The infection unfortunately persisted, and sadly, some months later, an above-knee amputation became necessary. The legal case, however, turned on not only whether the DAIR was negligent, but also whether the patient had given properly informed consent. Although the patient had accepted the surgeon’s best advice over more than a decade for her four previous operations, the Court found that the surgeon had been ‘… less than meticulous in his approach to consenting the claimant’ because the alternative of a definite two-stage procedure, with a deferred second arthroplasty, had not been discussed with the patient. The negligence part of the case failed on causation because the court found that, on the balance of probabilities (it was a civil case), while there was a defect in the informed consent, it had no impact on the patient’s unquestionably bad outcome.

The second case was from Scotland (McCullock and Others), and led to the UK Supreme Court being asked: “What test should be applied when assessing whether an alternative treatment is reasonable and requires to be discussed with the patient?” [17]. Again, the clinical case was not rare: the patient was a cardiology out-patient, but had died of pericarditis and cardiac tamponade outside of hospital. The Consultant Cardiologist was alleged not to have properly obtained informed consent for the treatments being used because of not discussing with the patient an alternative therapy, namely non-steroidal anti-inflammatory drugs (NSAIDs). The lower courts had found that even if efficacy with NSAIDs was very unlikely, that was irrelevant. The appeal failed.

In both of these cases, part of the Respondents’ reasoning was that patients would be confused with discussions about multiple, alternative, potential treatments, much of which might not be relevant, and which would potentially require a clinician to treat a patient against his/her best advice. Furthermore, a blunderbuss approach to such information could, to some extent, even absolve clinicians from having to identify what their best advice might be. As one solicitor (Vicki Swanton) has put it: “Patients should not be bombarded with information (much of which could be irrelevant) which was a real risk if all possible treatment modalities had to be shared with them at every stage of their treatment”. For example, in the first case (the infected knee implant), the patient had diabetes, was obese, had poor mobility and recurrent cellulitis in her legs. Thus, in her case, every general anaesthetic represented more than an average hazard, and thus, for her, compared with a two-stage procedure, there was a greater than average relative benefit of a single anaesthetic, if that was possible.

Clearly, there should be no return to ‘paternalistic’ models of healthcare. In his written opinion in Powell (and before demonstrating his clinical competency with a graph illustrating the time course of the patient’s C-reactive protein concentration in mg/L !), His Honour Judge Dias made this very clear:

“Let me emphasise at the outset that we live in an age where informed consent and patient autonomy are given great priority. Quite properly. To “consent” a patient is not a mere technicality, a box-ticking exercise for insurance or compliance purposes. Instead, it goes to the heart of legally authorising the substantial and at times severe interference with the patient’s right to bodily integrity. The court takes such questions very seriously.”,

and

“For in our modern law, “medical paternalism” no long rules. This is what is at stake in this case. This ethos permeates modern medical practice.”

Doubtless, the law is good at making post hoc decisions about proportionality of information for truly informed consent. However, the law is a crude tool for prospectively guiding what is reasonable in the unique circumstances of any particular patient. The solution is not simply a function of the presenting pathology and co-morbidities. The patient’s personality (in the psychological sense of the term), their own approach to life, cross-cultural aspects, and the time available for the patient to take independent advice, must all be taken into account. Lastly, these judgments would seem to have wider application, including being applicable to informed consent in clinical trials.


Happy New Year!