Top Texts from Prescrire International

I NT ERNAT I ONAL

SPECIAL EDITION

Funded by subscribers. No advertising, no grants, no shareholders

TOP TEXTS OF 2020

Specially selected to introduce you to this wholly independent monthly journal , written by and for healthcare professionals

Prescrire International • 2020 • SPECIAL EDITION

Illustration José David

Who are we? Prescrire is a French non-profit continuing education organisation, committed to better patient care Prescrire International , the English-language edition of La Revue Prescrire , provides independent information, by and for healthcare professionals.

INTERNATIONAL

engl i sh.prescri re .org

President of Association Mieux Prescrire: Pierre Chirac, pharmacist ASSOCIATION MIEUX PRESCRIRE  “To work, in all independence, in favour of quality healthcare, first and foremost in the interest of patients (...)” (Article 1 of the Association Mieux Prescrire bylaws). Prescrire International is a monthly journal of selected texts translated from the French journal Prescrire . Prescrire International is published 11 times a year by Association Mieux Prescrire, a non-profit continuing education organisation (N° 11 711 075) under the French law of 1901. EDITORIAL STAFF Members of the Prescrire Editorial Staff sign a yearly declaration of absence of conflicts of interest, in accordance with Prescrire’s “Non merci...” Charter. Members are free from any interest contrary to Association Mieux Prescrire’s objectives (the Charter and the Declaration are available online at english.prescrire.org). Publishing reliable reviews that are easy to use and adapted to readers’ needs depends upon complex editorial procedures, all initiated and overseen by the members of Prescrire’s Editorial Staff. Members of the Editorial Staff define editorial goals.They oversee the literature search, the writing and rewriting of texts, and the review by a panel of outside experts (medical specialists, methodologists, representatives of Prescrire’s subscriber base…).They organise internal and external quality control procedures, and edit the final copy. Every draft is submitted, before publication, to a large number of external reviewers. Copyright Prescrire www.prescrire.org Registered address: 83, bd Voltaire 75011 Paris France Tel.: (33)(0)1 49 23 72 80 E-mail: contact@prescrire.org Postal address: Prescrire - 83, bd Voltaire 75558 PARIS CEDEX 11 - FRANCE - Around 150 members of the Prescrire team - Around 100 external reviewers, experts in the subjects discussed and practising healthcare professionals For each issue, readers are provided with a list of the contributors to that issue. Around 250 people participate in the production and distribution of every issue of Prescrire , including:

them, patients – with the clear, compre- hensive and reliable information they need about drugs and therapeutic and diagnostic strategies. Association Mieux Prescrire publishes a monthly journal in French, and an edition in English 11 times a year. A non-profit organisation, Prescrire is wholly financed by its subscribers, and accepts no advertising or other outside support. Reliable and relevant content Prescrire has the editorial and research capabilities necessary to ensure the accuracy of its reviews. Prescrire ’s editors are healthcare professionals, specially trained in Prescrire ’s editorial methods and free from conflicts of interest. Exacting quality-control procedures are applied to all editorial content.

The Association Mieux Prescrire (AMP), a non-profit organisation registered under the French law of 1901, manages all of Prescrire ’s programmes and publications. The AMP is structured so as to be free of any influence from pharma- ceutical companies or healthcare institutions. Independent information Prescrire’ s purpose is stated in Article 1 of the AMP bylaws: “To work, in all independence, in favour of quality health- care, first and foremost in the interest of patients (...).” Prescrire offers continuing education and professional practice improvement programmes specifically adapted to the needs of healthcare professionals. For nearly 40 years, Prescrire has pro- vided healthcare professionals – and via

Subscribers can download any of the thousands of detailed reviews produced b y Prescrire’s multidisciplinary team of healthcare professionals, and all including complete notes and references. All are available instantly , via an online table of contents going back to 1992, or via a search function.

SUBSCRIBE NOW FOR FULL ACCESS TO ONLINE CONTENT

Prescrire is a founder and member of ISDB, a network of drug bulletins that are independent of drug companies, which aims to promote international exchange of quality information on drugs and therapeutics. Website: www.isdbweb.org

Reliable, rigorously independent information on treatments and healthcare strategies, to enable fully informed decision-making. No grants, no advertising. No shareholders, no sponsors. Prescrire is funded by its subscribers.

Bulls Market - Group/TPI

Editorial  New drugs: the right to know TOP TEXTS OF 2020 SPECIAL EDITION How we work Prescrire ’s complex, collective editing process has been fine-tuned over the years NEW PRODUCTS

2

Pembrolizumab ( keytruda ° ) with cytotoxic drugs in metastatic lung cancer  A useful first-line combination when PD-L1 expres- sion is low Opicapone ( ongentys ° ) and Parkinson’s disease with motor fluctuations  A single daily dose, but more dyskinesia Dengue vaccine ( dengvaxia ° ) Not for large-scale use Denosumab ( prolia ° ) and steroid-induced osteoporosis EMA turns a blind eye to denosumab’s lack of proven clinical efficacy

3-4

Prescrire ’s reviews, written by specially trained health professionals, are based on an exhaustive search of the literature, and undergo scrutiny by a large panel of outside reviewers plus rigorous quality-control procedures.

5-6

7

Editors free from conflicts of interest The absence of any direct or indirect financial links to the pharmaceutical industry is an absolute requirement to be a member of the Prescrire team. Any such link would be cause for dismissal from the Prescrire Editorial Staff. A vast network of reviewers Once they are in an advanced stage of editing and quality control, all draft reviews are sent to outside reviewers (10 to 40 reviewers read each draft at this stage). These reviewers critique each article in terms of content, style, relevance, newsworthiness, presenta- tion of arguments and usefulness in daily practice. The reviewers are spe- cialists in the subject being discussed, methodologists, or healthcare profes- sionals representative of Prescrire ’s readers (and chosen from their midst).

All the reviews published in la Revue Prescrire (aside from a few clearly labelled exceptions, such as readers’ letters) are written by Prescrire ’s Edi- torial Staff. Prescrire does not publish unsolicited manuscripts from outside contributors. The production of reviews draws upon a wide range of skills, all exer- cised under the supervision of Prescrire ’s editors. This team approach is reflected in the collective byline “©Prescrire”. Written and edited by healthcare professionals Most of Prescrire ’s editors are physicians (both in individual prac- tice or on hospital staff), pharma- cists (working in pharmacies or hos- pitals), nurses and dentists. A few are economists or journalists with specific expertise in the area of health- care. All Prescrire editors have received extensive in-house training in Prescrire’ s editorial production process.

8

9

ADVERSE EFFECTS

High thrombotic risk and antiphospholipid antibodies Avoid direct oral anticoagulants Antiepileptics and pregnancy Potential long-term effects in children

10

11

Oral peanut desensitisation Excessive adverse effects

12

SSRI antidepressants  Persistent sexual dysfunction

12

REVIEWS

Dulaglutide and type 2 diabetes No reduction in all-cause mortality or cardiovascular mortality Antispasmodics during pregnancy In brief

13

14

Follow Prescrire International on Facebook andTwitter

OUTLOOK

Gifts to doctors wield undue influence in France

15

Keep up with the latest from Prescrire ’s editors, and support our values of independence and quality information, first and foremost in the interest of patients. Follow us

Drugs in 2019: a brief review

16-17

Towards better patient care: drugs to avoid in 2020

18-20

P rescrire I nternational S pecial E dition 2020 • P age 1

New drugs: the right to know

According to its officials, the role of the European Medicines Agency (EMA) is to ensure that patients have timely access “to medicines that are safe, effective and of suitable quality, as well as the information needed to use those medicines and make informed choices about their treatment” (1). Does the EMA really fulfil these missions? Since the second decade of the 21 st century, more and more publications have been critical of marketing authorisations (MAs), both in the United States and Europe. These publications show that a very significant proportion of drugs are introduced to the market in the absence of any proof of tangible clinical benefit for patients (2-5). For example, in 2019, a study showed that more than half of all drugs entering the German market had no demonstrated added benefit over standard care (2). Another study showed that the majority of drugs authorised in Europe between 2011 and 2018 through the accelerated MA procedure were evaluated on the basis of surrogate endpoints without any correlation with clinical outcomes (3). A study of cancer drugs approved by the EMA between 2014 and 2016 showed that half of the randomised clinical trials supporting the MAs were probably biased in their design, their conduct or the analysis of their results (4). According to EMA’s medical director, patients have to accept a significant level of uncertainty with regard to new drugs (6). EMA officials recognise that “more emphasis should be placed on contextualizing the effect of new medicines and on being more explicit about negative, neutral or positive added benefit where possible in relevant patient subgroups” (7). These officials had already recognised in 2018 that the added therapeutic benefit of a new drug could be negative (8). Of course, uncertainties exist at the time of marketing authorisation, but what is the EMA doing to explain these uncertainties to patients? What is the EMA doing to make healthcare professionals and patients aware that new drugs authorised in Europe can have a “negative” added therapeutic value, i.e. represent a step backwards? We look forward to the EMA explaining and providing details in drug information documents, particularly in the summary of product characteristics and patient information leaflets, concerning the weakness of the clinical evidence supporting the marketing authorisation, and the magnitude of the resulting uncertainties regarding the real therapeutic benefit of drugs. This would enable healthcare professionals and patients to make more informed choices, just as the EMA is claiming in its mission statement. Prescrire ▶ Translated from Rev Prescrire January 2020 Volume 40 N° 435 • Page 56  References  1- European Medicines Agency “EMA Regulatory Science to 2025. Strategic reflection 2018”: 60 pages.  2- Wieseler B “New drugs: where did we go wrong and what can we do better?” BMJ 2019; 366: I4340: 8 pages.  3- Schuster Bruce C et al. “The use of validated and nonvalidated surrogate endpoints in two European MedicinesAgency expedited approval pathways: a cross-sectional study of products authorised 2011-2018” PloSMed 2019; 16 (9): e1002873: 30 pages.  4- Naci H et al. “Design characteristics, risk of bias, and reporting of randomised controlled trials supporting approvals of cancer drugs by European Medicines Agency, 2014-16: cross sectional analysis” BMJ 2019; 366: l5221: 17 pages.  5- Del Paggio JC andTannock F “The fragility of phase 3 trials supporting FDA - approved anticancer medicines: a retrospective analysis” Lancet Oncol 2019: 20 (8): 1065-1069.  6- Eichler HG et al. “From adaptive licensing to adaptive pathways: delivering a flexible life-span approach to bring new drugs to patients” Clin PharmacolTher 2015; 97 (3): 234-246.  7- Eichler HGet al.“Added therapeutic benefit and drug licensing” Nat Rev DrugDiscov 2019; 18: 651-652.  8- Eichler HG“Adaptive pathways. Reply to Prof. SilvioGarattini, etc.” 16 June 2016 EMA/365120/2016: 8 pages.

EDITORIAL

Prescrire Int • April 2020

P age 2 • P rescrire I nternational S pecial E dition 2020

NEW PRODUCTS

pembrolizumab ( keytruda °) with cytotoxic drugs in metastatic lung cancer A useful first-line combination when PD-L1 expression is low

 OFFERS AN ADVANTAGE  In two clinical trials, adding pembrolizumab to first-line platinum-based chemotherapy in patients with metastatic non-small cell lung cancer extended survival by a few months. The addition of pembrolizumab increases the incidence of certain adverse effects. If over half of the patient’s tumour cells express the PD-L1 protein, the addition of platinum-based chemotherapy to pembrolizumab has not been shown to provide any clinical benefit. If fewer than half of the tumour cells express PD-L1, and platinum-containing chemother- apy is chosen as first-line treatment, pem- brolizumab add-on therapy should be offered to patients who are sufficiently well to toler- ate the added harms. KEYTRUDA° - pembrolizumab concentrate for solution for intravenous infusion ■ immunostimulant; anti-PD-1 ■ New indications: first-line treatment of adults with meta- static non-small cell lung carcinoma (NSCLC): in combination with pemetrexed and platinum chemotherapy for non- squamous NSCLC with no EGFR or ALK positive mutations; in combination with carboplatin and paclitaxel for squamous NSCLC. [EU centralised procedure] The first-choice treatment for patients with meta- static non-small cell lung cancer who have not yet received treatment at this stage of the disease is often platinum-based chemotherapy in combination with another cytotoxic drug such as paclitaxel or pemetrexed . Pemetrexed appears less effective than other cytotoxic drugs against tumours classified as squamous (1-3). But when at least half of the tumour cells express the PD-L1 protein (which binds to the PD-1 receptor), monotherapy with pembrolizumab (an immuno- stimulatory anti-PD-1 monoclonal antibody) is a first-choice option for patients who have not yet received treatment at this stage of the disease. In this situation, two clinical trials versus platinum- based chemotherapy showed that pembrolizumab monotherapy extends median survival by about Compare before deciding

1 year more than chemotherapy, without increasing the incidence of serious adverse effects (1).

What’s new?

Pembrolizumab (Keytruda°, Merck Sharp & Dohme) has been authorised as add-on therapy to certain platinum-based chemotherapy regimens for first- line treatment of metastatic non-small cell lung cancer, irrespective of the proportion of tumour cells expressing PD-L1 (4). What does the addition of platinum-based chemo- therapy to pembrolizumab offer patients, compared with pembrolizumab alone, when at least half of their tumour cells express PD-L1? What does the addition of pembrolizumab to platinum-based chemotherapy offer patients when fewer than half of their tumour cells express PD-L1? What are the adverse effects of the combination of pembrolizu- mab and chemotherapy? Prolonged median survival in two double-blind clinical trials. Evaluation of pembrolizumab + platinum-based chemotherapy in patients with metastatic non-small cell lung cancer is based on two double-blind randomised trials of pembrolizumab versus placebo: one trial in 616 patients with non- squamous disease, added to the combination of a platinumcompound + pemetrexed (the“Keynote-189” trial); and one trial in 559 patients with squamous cell carcinoma, added to the combination carboplatin + paclitaxel (the “Keynote-407” trial) (4-7). In both trials, all the patients were considered well or in reasonably good general condition despite their cancer. In about 30% of cases, at least half of the patient’s tumour cells expressed PD-L1 (5,6). The results available are from protocol-planned interim analyses (5,6). As about 30% of patients in the placebo groups received pembrolizumab when their cancer worsened, the effects of pembrolizu- mab may be greater than suggested by the differ- ences observed between the groups (5,6). We found no studies that compared chemotherapy + pembrolizumab versus pembrolizumab alone in patients in whom at least half of the tumour cells expressed PD-L1. Median survival extended by a few months com- pared with placebo. In an analysis of the Keynote- 189 trial, performed after a median follow-up of about 10 months, the 6-month survival rate in the pembrolizumab group was 85%, versus 72% in the placebo group (p<0.00001) (5). The abstract of a

P rescrire I nternational S pecial E dition 2020 • P age 3

NEW PRODUCTS

Assessment elsewhere

conference presentation from May 2019 reported that after half of the patients had been followed up for at least 19 months, the estimated median sur- vival in the pembrolizumab group was 22 months, versus 11months in the placebo group (p<0.00001) (8). In an analysis of the Keynote-407 trial, conducted after a median follow-up of about 8 months and the death of 205 patients, the estimated median sur- vival was 16 months in the pembrolizumab group, versus 11 months in the placebo group (p=0.0008) (6). In both trials, longer survival was observed in particular in the subgroups of patients in whom fewer than half of the tumour cells expressed PD-L1 (5,6). It has not been shown whether pem- brolizumab + chemotherapy is more effective than pembrolizumab alone in patients in whom more than half of the tumour cells express PD-L1, as this has not been evaluated in a head-to-head trial. Increased incidence of certain adverse effects. The known adverse effects of pembrolizumab are main- ly of immunological origin, with possible effects on multiple organs: pneumonitis or interstitial lung disease, myocarditis, colitis, stomatitis, gastritis, pancreatitis, hepatitis, nephritis, uveitis, vasculitis, severe cutaneous adverse reactions, encephalitis and neuropathy including Guillain-Barré syndrome, endocrinopathies including thyroid disease and diabetes, myasthenia gravis, and haemolytic an- aemia. Pembrolizumab exposes to gastrointestinal disorders, fatigue, musculoskeletal pain, allergic reactions including anaphylactic reactions, fever, and alopecia (4,9). No previously unknown adverse effects of pem- brolizumab were identified in the clinical trials de- scribed above (5,6) In these trials, about 26% of patients in the pem- brolizumab groups versus about 20% in the place- bo groups had a serious adverse effect considered treatment-related by the investigators (who did not know which treatment the patient had received). And about 20% of patients versus 7% discontinued treatment due to an adverse effect (5,6). The addition of pembrolizumab appeared to in- crease the incidence of certain adverse effects. For example, patients in the pembrolizumab group of the trial in non-squamous disease were more like- ly to have diarrhoea (reported in 31% of patients versus 21% in the placebo group), rash (25% ver- sus 17%), neutropenia (48% versus 41%) and in- creased serum creatinine (37% versus 25%). In the trial in patients with squamous cell carcinoma, those in the pembrolizumab group were more likely to have alopecia (47% versus 36%) and peripheral neuropathy (31% versus 25%) (5,6,10).

The clinical evaluation data on pembrolizumab as add-on therapy to first-line platinum-based chemo- therapy in certain types of metastatic non-small cell lung cancer were analysed by the Transparency Committee of the French National Authority for Health (HAS). Its main conclusion is that the addition of pembrolizumab represents a moderate therapeut- ic advance, provided the patient’s condition is good. The committee highlighted the lack of comparative trials of pembrolizumab + chemotherapy versus pembrolizumab alone in patients in whom more than half of tumour cells express PD-L1 (7). For patients with metastatic non-small cell lung cancer, if fewer than half of their tumour cells ex- press PD-L1, and platinum-based chemotherapy is chosen as first-line treatment, add-on therapy with pembrolizumab should be offered if the patient is sufficiently well to tolerate the increased incidence of certain adverse effects and is willing to do so. If more than half of the patient’s tumour cells express PD-L1, the addition of chemotherapy to pembrolizumab has not been shown to provide any clinical benefit as of late 2019, but this combination is likely to increase the risk of adverse effects. ©Prescrire ▶ Translated from Rev Prescrire December 2019 Volume 39 N° 434 • Pages 892-894 In practice Literature search up to 9 October 2019 In response toour request for information,MSDprovided us with no documentation on its product. 1- Prescrire Editorial Staff “Metastatic lung cancer. Pembrolizumab as first-line therapy: prolonged survival in some patients” Prescrire Int 2019; 28 (210): 302. 2- PrescrireRédaction“Pémétrexed: limiter lesatteinteshématologiques avec l’acide folique” Rev Prescrire 2017; 37 (408): 745-746. 3- Prescrire Editorial Staff“Pemetrexed. Untreated non-small cell lung cancer: just another cytotoxic drug” Prescrire Int 2009; 18 (101): 114. 4- EuropeanCommission“SPC-Keytruda” 26August 2019: 105 pages. 5- EMA - CHMP“Public assessment report for Keytruda. EMEA/H/C/ 003820/II/0043” 26 July 2018: 89 pages. 6- EMA - CHMP“Public assessment report for Keytruda. EMEA/H/C/ 003820/II/0060” 31 January 2019: 103 pages. 7- HAS - Commission de laTransparence“Avis-Keytruda” 20 February 2019 and 9 October 2019: 28 pages and 29 pages. 8- Gadgeel SMet al.“Keynote-189: updatedOS and progression after the next line of therapy (PFS2) with pembrolizumab (pembro) plus chemo with pemetrexed and platinum vs placebo plus chemo for metastaticnonsquamousNSCLC” JClinOncol 2019; 37 (15suppl):9013. 9- Prescrire Rédaction “Anti-PCD-1 ou anti-PCD-L1: nivolumab, etc.” Interactions Médicamenteuses Prescrire 2019. 10- US FDA“Full prescribing information-Keytruda” September 2019: 80 pages.

Prescrire Int • February 2020

P age 4 • P rescrire I nternational S pecial E dition 2020

NEW PRODUCTS

opicapone ( ongentys °) and Parkinson’s disease with motor fluctuations A single daily dose, but more dyskinesia

 NOTHING NEW  In a comparative, double-blind, randomised trial in 600 patients suffering from Parkinson’s disease and affected by motor fluctuations, opicapone taken once daily has not been shown to be more effective in reducing the duration of “off” periods than entacapone taken with each dose of levodopa .The adverse effect profile of opicapone seems to be close to that of entacapone , but is less well known. Opicapone carries a greater risk of dyskin­ esia. In practice, opicapone does not represent a real advance.

The onset of motor fluctuations on levodopa should initially be managed by optimising dopamin­ ergic therapy, e.g. dividing the daily dose into several administrations, adjusting the timing of the doses, or even increasing the daily dose or resort- ing to an extended-release formulation. These measures are sometimes difficult to implement, particularly in patients with swallowing difficul- ties (1,2). In cases where these measures are insufficient, addition of entacapone is an option. By inhibiting peripheral catechol-O-methyltransferase (COMT), entacapone reduces the breakdown of levodopa , which leads to an increase in the quantity of levo- dopa available in the brain. Fixed-dose combinations of entacapone + levodopa + carbidopa reduce the number of tablets to be taken, but the ratio of levodopa to carbidopa doses is inadequate, expos- ing patients to an increased risk of dyskinesia and nausea ( b )(3-5). Tolcapone , another COMT inhibitor, carries a risk of liver damage, which is sometimes fatal, making its harm-benefit balance unfavourable. Liver dam- age has rarely been reported with entacapone (5). Opicapone (Ongentys°, Bial Portela) is a COMT inhibitor, like entacapone. It has been authorised in the European Union for use in patients with Parkin- son’s disease who are affected by motor fluctuations despite dopaminergic therapy (6). When a COMT inhibitor is considered in this situ­ ation, is addition of opicapone more effective in improving motor symptoms than addition of enta- capone taken separately with each dose of levo- dopa ? What are the adverse effects of opicapone , in particular on the liver? No more effective than entacapone on the duration of “off” periods. Clinical evaluation of opicapone in patients with Parkinson’s disease only includes one clinical trial that compared it to entacapone . What’s new?

ONGENTYS° - opicapone capsules • 50 mg of opicapone per capsule ■ antiparkinsonian drug; peripheral catechol-O- methyltransferase (COMT) inhibitor ■ Indication: in combination with dopaminergic therapy ( levodopa + peripheral dopa-decarboxylase inhibitor) in adult patients with Parkinson’s disease affected by end-of- dose motor fluctuations. [EU centralised procedure]

Compare before deciding

Parkinson’s disease is a progressive central nervous system disorder. It manifests mainly through motor symptoms linked to progressive degeneration of the cerebral dopaminergic neurons: tremor at rest, slowness and poverty of movement, and rigidity (or stiffness) (1,2). In patients with Parkinson’s disease, treatment is aimed at relieving disease symptoms and maintain- ing satisfactory independence. Drugs are justified when the functional impairment constitutes a handicap. The first choice is then dopaminergic ther- apy that combines levodopa (a dopamine precursor) with a peripheral dopa-decarboxylase inhibitor ( car- bidopa or benserazide ) in order to reduce the gas- trointestinal and cardiovascular adverse effects of levodopa . All patients initially respond to dopami- nergic therapy. However, after 4 to 6 years of use, about 40% of patients treated with dopaminergic therapy experience motor fluctuations (end-of-dose akinesia, and “on-off” phenomena) or dyskinesia, often mid-dose.This proportion is around 90% after 9 years ( a )(1-4).

a- According to some reviewers of the draft text, the“on-off” phenomena aremore troublesome for patients thandyskin­ esia. b- In France, the immediate-release tablets containing levodopa + carbidopa have a dose ratioof 10:1. Infixed-dose combinations of levodopa + carbidopa + entacapone the dose ratio of levodopa to carbidopa is 4:1 (ref 10).

P rescrire I nternational S pecial E dition 2020 • P age 5

NEW PRODUCTS

This so-called non-inferiority trial included 600 pa- tients treated with dopaminergic therapy and ex- periencing end-of-dose motor fluctuations (6-8). After randomisation, in addition to the dopamin- ergic therapy, they received in double-blind manner either opicapone (5 mg per day, 25 mg per day, or 50 mg per day), or entacapone , or placebo.The dose of levodopa could be adjusted during the first two weeks after randomisation, but not thereafter (6). This article only reports the results obtained with opicapone 50 mg per day, the only dose recom- mended in the marketing authorisation. At baseline, the patients had on average received dopaminergic therapy for about 5 years and had been suffering from motor fluctuations for about 2 years.The average duration of “off” periods was about 6 hours per day in the different groups. Most of the patients were receiving other antiparkinsonian drugs (most often a dopaminergic agonist), with no notable difference between the groups (6). In this trial, the primary outcome measure was the mean reduction in the duration of “off” periods after 14 to 15 weeks of treatment, compared to baseline. This reduction was 117 minutes in the opicapone group, which was close to that in the entacapone group (96 minutes) and statistically significantly greater than that in the placebo group (56 minutes) (6). Dyskinesia twice as prevalent as with entacapone. The predictable adverse effect profile of opicapone is that of the other COMT inhibitors, mainly related to the increased dopaminergic action of levodopa : gastrointestinal disorders (including chronic diarrhoea and colitis), orthostatic hypotension, mental health problems (including compulsive disorders, insomnia, hallucinations and abnormal dreams), excessive drowsiness and suddenly falling asleep during the daytime, plus rhabdomyolysis and liver damage (especially with tolcapone ) (5,9).There is a possible increased risk of prostate cancer with entacapone . In the double-blind randomised trial versus enta- capone , the incidence of adverse effects was about 33% in the opicapone 50 mg per day group and in the entacapone group, but there was an increase in dyskinesia with opicapone : 15% versus 7% with entacapone , versus 4% with placebo. The other adverse effects more frequently reported in the entacapone and opicapone groups than in the pla- cebo group were constipation, nausea, insomnia and dizziness (7). Taking into account all the trials that evaluated opicapone , particularly those versus placebo, there was an indication of a possible increase in ischaemic heart disorders. Abnormal liver function has also been reported in a few patients treated with opica- pone (6). Beware of hypertensive crises when used in com- bination with a monoamine oxidase inhibitor (MAOI). As with the other COMT inhibitors, the combination of opicapone with a monoamine oxi- dase inhibitor (MAOI) leads to an inhibition of

catecholamine metabolism, exposing patients to the risk of hypertensive crises. Combination of opicapone with an MAOI (such as selegiline , an- other antiparkinsonian drug), a tricyclic antidepres- sant, a drug that has an inhibitory action on nor- adrenaline reuptake (such as the antidepressant venlafaxine ) or a drug metabolised by COMT (such as the antidepressant paroxetine ) should be avoid- ed, or if not, requires close clinical monitoring (5,9). Opicapone and its metabolites inhibit, in particu­ lar, the cytochrome P450 isoenzyme CYP 2C8, creating a risk of overdose when combined with substrates of this isoenzyme, such as repaglinide (a hypoglycaemic drug) (6,9). Convenience: just one administration per day, but cannot be taken simultaneously with levodopa. Opicapone is taken only once daily, but it must be taken at least one hour before or after levodopa . Entacapone, on the other hand, is taken up to 10 times per day, along with each dose of levodopa + peripheral dopa-decarboxylase inhibitor (9). In patients with Parkinson’s disease suffering from motor fluctuations despite optimised dopaminergic therapy, opicapone has not been shown to be more effective than entacapone in reducing the duration of “off” periods. Opicapone increases the risk of dyskinesia, and its adverse effect profile is less well known than that of entacapone . As of 2019, there is no reason to choose opicapone over entacapone , despite the significant reduction in the number of daily doses that need to be taken. ©Prescrire ▶ Translated from Rev Prescrire September 2019 Volume 39 N° 431 • Pages 652-654 In practice Literature search up to 8 July 2019 In response to our request for information, Bial Portela provided uswith administrative documents and published articles . 1- Prescrire Rédaction “Maladie de Parkinson: traitement initial des troubles moteurs” Premiers Choix Prescrire, updated January 2018: 4 pages. 2- Prescrire Rédaction “Traitement de la maladie de Parkinson. Deux- ième partie. Réduire les fluctuations motrices sous lévodopa” Rev Prescrire 2011; 31 (330): 273-279. 3- Prescrire Rédaction“Entacapone: une autre option dans lamaladie de Parkinson” Rev Prescrire 2013; 33 (353): 187-188. 4- Prescrire Rédaction “Stalevo°, un nième dosage: ça suffit!” Rev Prescrire 2013; 33 (359): 660. 5- Prescrire Rédaction“Patients parkinsoniens” InteractionsMédica- menteuses Prescrire 2019. 6- EMA - CHMP “Public assessment report for Ongentys. EMEA/ H/C/002790/0000” 28 April 2016: 140 pages. 7- HAS - Commission de la transparence “Avis-Ongentys 50 mg” 23 January 2019: 27 pages. 8- Ferreira JJ et al. “Opicapone as an adjunct to levodopa in patients with Parkinson’s disease and end-of-dose motor fluctuations: a ran- domised, double-blind, controlled trial” Lancet Neurol 2016; 15 (2): 154-165. 9- European Commission “SPC-Ongentys” 13 March 2019 + “SPC-­ Comtan” 19 April 2018: 37 pages. 10- Prescrire Editorial Staff “Levodopa + carbidopa + entacapone, Parkinson’s disease: amodest effect” Prescrire Int 2005; 14 (76): 51-54

Prescrire Int • January 2020

P age 6 • P rescrire I nternational S pecial E dition 2020

NEW PRODUCTS

dengue vaccine ( dengvaxia °) Not for large-scale use

Abstract

● As the dengue vaccine is a live attenuated vac- cine, pregnant women and immunocompromised patients are at risk of developing serious infec- tion with the viruses contained in the vaccine. NOTHING NEW  The dengue vaccine reduced the incidence of dengue, including severe forms, in children living in endemic areas. However, this vac- cine is unsuitable for mass immunisation in endemic areas as of 2019, because it can increase the risk of severe dengue in persons not previously infected with dengue virus. Yet no test suitable for routine screening and proven to possess the required performance characteristics is available to identify these children. In practice, the only population in which the harm-benefit balance of this dengue vaccine appears favourable is chil- dren for whom there is documented evidence of positive serological testing at the time of a prior infection. Its efficacy appears to wane over time and the benefit of a booster dose has not been evaluated. As of 2019, the prevention of dengue relies above all on vector control in endemic areas, and person- al protection against mosquito bites. DENGVAXIA° - dengue vaccine powder and solvent for suspension for subcutaneous injection ■ dengue tetravalent vaccine ■ Indication: “ prevention of dengue disease caused by den- gue virus serotypes 1, 2, 3 and 4 in individuals 9 to 45 years of age with prior dengue virus infection and living in endem- ic areas ”. [EU centralised procedure] ■ Dosage: “ 3 injections of one reconstituted dose (0.5 ml) to be administered at 6-month intervals ”.

● Dengue is an infection caused by a virus with 4 known serotypes.The infection is often asymp- tomatic and generally self-limiting.When infect- ed for a second time, a patient’s risk of severe and potentially fatal dengue is increased. As of late 2019, dengue prevention is mainly based on control of the mosquitoes that transmit the virus in endemic areas (tropical or subtropical regions) and personal protection against mos- quito bites. ● A vaccine containing live attenuated viruses expressing antigens of the 4 known dengue virus serotypes (Dengvaxia°, Sanofi Pasteur) has been granted marketing authorisation in the European Union. It is authorised for use in persons aged 9 years to 45 years who have pre- viously been infected with the virus and live in an endemic area. ● The dengue vaccine has not been evaluated in persons living in non-endemic regions or in travellers visiting an endemic region. In three randomised placebo-controlled trials in about 35 000 children aged 2 to 16 years living in endem- ic areas, the vaccine was effective in reducing the incidence of symptomatic dengue, including severe forms, during the two years following the first injection. Its efficacy increased with age. Preliminary data from long-term follow-up sug- gest that its efficacy diminishes over time. ● An increased incidence of severe dengue was observed in vaccinated children aged 2 to 5 years, and in vaccinated children who had not previously been infected with one of the den- gue viruses.Yet no test that is sufficiently sim- ple to perform and has the required proven per- formance characteristics for routine identification of prior infection is available as of 2019. ● No trials have evaluated the effect of the vac- cine on clinical outcome measures in adults aged 18 to 45 years. ● About half of the patients included in trials had an injection site reaction. The systemic adverse effects were those generally observed with vaccines: headache, fatigue, myalgia, mal- aise, and allergic reactions, including anaphyl- actic reactions.

©Prescrire ▶ Excerpt from Rev Prescrire November 2019 Volume 39 N° 433 • Page 810

Prescrire Int • March 2019

P rescrire I nternational S pecial E dition 2020 • P age 7

NEW PRODUCTS

denosumab ( prolia °) and steroid-induced osteoporosis

NOT ACCEPTABLE Denosumab has not been shown to prevent clinical fractures, but it has many adverse effects that can be severe or even fatal, includ- ing deep infections, hypocalcaemia, osteo- necrosis and multiple vertebral fractures after discontinuation. PROLIA° - denosumab solution for subcutaneous injection • 60 mg of denosumab per prefilled syringe in 1 ml of solution ■ anti-RANKL monoclonal antibody ■ New indication : “ bone loss associated with long-term systemic glucocorticoid therapy in adult patients at increased risk of fracture ”. [EU centralised procedure] Prolonged systemic corticosteroid therapy leads to bone loss and an increased risk of fracture. Fracture prevention in patients receiving long-term cortico- steroid therapy relies mainly on non-pharmacological measures (fall prevention, regular exercise, adequate dietary intake of calcium and vitamin D), while using the lowest effective dose of the corticosteroid for the shortest possible duration. Bisphosphonates have not been shown to prevent symptomatic fractures in this situation (1,2). Denosumab (Prolia°, Amgen) is a monoclonal antibody that binds to a cytokine called RANKL that stimulates osteoclast activity and plays a role in immunity. Denosumab has an unfavourable harm­ benefit balance in the prevention of osteo- porotic fractures. It has also been authorised in the prevention of osteoporosis induced by prolonged corticosteroid therapy (3-5). Denosumab has mainly been evaluated in this situation in a double-blind randomised trial in 795 adults (median age 63 years) who had been taking prednisone at a daily dose of at least 7.5 mg, for over 3 months in two-thirds of cases.The patients were randomised to receive either denosumab or the bisphosphonate risedronic acid for 2 years, in conjunction with calcium and vitamin D supplemen- tation. After one year of treatment, the increase in bone mineral density (the primary endpoint) was greater in the denosumab group than in the rise- dronic acid group (p<0.001). This effect persisted after 2 years of treatment (4). However, a similar proportion of patients in the two groups sustained symptomatic fractures (about 5% after one year of treatment) (4).

Denosumab has many adverse effects, which can be severe or even fatal: back, muscle and bone pain; immunosuppressive effects resulting in deep infec- tions including endocarditis and skin infections; hypocalcaemia; osteonecrosis of the jaw or external auditory canal; cataracts; cancer; hypersensitivity reactions; pancreatitis; cardiovascular disorders; autoimmune diseases; alopecia; and multiple ver- tebral fractures after discontinuation (3,5-7). No previously unknown adverse effects were reported during the trial discussed above (4). The incidence of serious adverse effects after 2 years of treatment was 24% in both groups (4). ©Prescrire ▶ Translated from Rev Prescrire January 2020 Volume 40 N° 435 • Pages 11-12 Literature search up to 12 November 2019 In response to our request for information,Amgen provid- ed us with no documentation on its product. 1- Prescrire Editorial Staff “Corticosteroids: no drug prevention of fractures needed” Prescrire Int 2009; 18 (102): 175. 2- AllenCS et al.“Bisphosphonates for steroid-induced osteoporosis” (CochraneReview) (last update 2016). In:“TheCochrane Library” John Wiley and Sons, Chichester 2016; issue 10: 108 pages. 3- Prescrire Editorial Staff “Denosumab and male osteoporosis. Do not use in men (or in women)” Prescrire Int 2016; 25 (168): 36. 4- EMA-CHMP“PublicassessmentreportforProlia.EMEA/H/C/001120/ II/0068” 26 April 2018: 74 pages. 5- Prescrire Rédaction“dénosumab” InteractionsMédicamenteuses Prescrire 2020. 6- Prescrire Editorial Staff “Denosumab: immune dysfunction” Prescrire Int 2018; 27 (198): 268-269. 7- EMA - PRAC“Minutes of the meeting on 08-11 April 2019” 16May 2019: 94 pages.

Prescrire Int • April 2020

P age 8 • P rescrire I nternational S pecial E dition 2020

NEW PRODUCTS

EDITORS’ OPINION

EMA turns a blind eye to denosumab's lack of proven clinical efficacy

I n the absence of a better alternative, one can of course use a drug known to carry a risk of serious adverse effects, provided that this is a fully-informed decision, and that the clinical benefits are significant and very clearly demonstrated. In other words, that the harm-benefit balance is favourable in the clin- ical setting in question. Denosumab at a dosage of 60 mg has been mar- keted in France since 2012. It carries a risk of nu- merous adverse effects, in particular infections, cancer, hypersensitivity reactions, osteonecrosis of the jaw and external auditory canal, multiple ver- tebral fractures after discontinuation of the drug, serious, even fatal, hypocalcaemia, and auto-immune disorders. The European Medicines Agency (EMA) issued a favourable opinion regarding extension of the au- thorisation for this drug to prevention of osteo- porosis caused by long-term corticosteroid therapy. As a result of this opinion, this indication was added to its marketing authorisation (MA) (see “Denosumab (Prolia) and steroid-induced osteopor­

osis” p. 8). In light of its already extensive profile of known adverse effects, one would expect that its evaluation in the prevention of corticosteroid-induced osteoporosis would be particularly robust and based on clinical criteria that are useful for patients. How- ever, detailed analysis of the evaluation shows that it falls short of the mark: only a single trial has as- sessed the effect of denosumab , with a radiological criterion as the primary outcome measure and no proof of clinical efficacy. How is it possible that the EMA gives more weight to hypothetical clinical benefits than to serious, well-recognised clinical adverse effects?What kind of blinkers is the EMA wearing that prevent it from seeing the patients lying by the side of the road? (see “New drugs: the right to know” p. 2). Prescrire ▶ Translated from Rev Prescrire January 2020 Volume 40 N° 435 • Page 4

Prescrire Int • April 2020

PRESCRIRE’S RATINGS Our judgement is based on the therapeutic advance of the new product. It considers not only the inherent value of each product in terms of its harm-benefit balance, but also its advantages and disadvantages relative to existing products available in France. Note that the relative value of new products can vary from one country to another.

Quality of information from pharmaceutical companies In response to our systematic requests

 BRAVO  The product is a major therapeutic advance in an area where previously no treatment was available.  A REAL ADVANCE  The product is an important therapeutic advance but has certain limitations.  OFFERS AN ADVANTAGE  The product has some value but does not fundamentally change the present therapeutic practice.  POSSIBLY HELPFUL  The product has minimal additional value, and should not change prescribing habits except in rare circumstances.

 NOTHING NEW  The product is a new substance but with no evidence that it has more clinical value than other substances of the same group. It can be a me-too or a near me-too.

Company provided detailed information including unpublished

data and packaging items.

Company provided information limited to published administrative data or packaging items.

 NOT ACCEPTABLE  Product without evident benefit but with potential or real disadvantages.

Company provided minimal information, mainly administrative

 JUDGEMENT RESERVED  The editors postpone their rating until better data and a more thorough evaluation of the drug are available.

and packaging items.

Company provided no information.

P rescrire I nternational S pecial E dition 2020 • P age 9

ADVERSE EFFECTS

High thrombotic risk and antiphospholipid antibodies: avoid direct oral anticoagulants

 In practice   Direct oral anticoagulants are some- times an alternative to vitamin K antagonists. However, given the increased risk of thrombosis in patients with antiphospholipid antibodies, it makes sense to avoid direct oral anticoagulants in this situation, and to consider switching to warfarin . Warfarin is the first-choice oral anticoagulant for patients at high risk of thrombosis, but also for those at highest risk of bleeding or drug inter- actions, because the dose can be adjusted on the basis of INR (international normalised ratio) meas- urements (3). These are situations faced by many more patients than those with antiphospholipid antibodies. ©Prescrire ▶ Translated from Rev Prescrire October 2019 Volume 39 N° 432 • Pages 736-737 a- The direct oral anticoagulants authorised in the European Union are the direct thrombin inhibitor dabigatran and the factor Xa inhibitors (or“xabans”) apixaban, edoxaban, and rivaroxaban.  Sources 1- ANSM “L’apixaban (Eliquis°), le dabigatran etexilate (Pradaxa°), l’edoxaban (Lixiana°/Roteas°) et le rivaroxaban (Xarelto°) ne sont pas recommandés chez les patients présentant un syndrome des antiphospholipides (SAPL) en raison d’une possible augmentation du risque de récidive d’évènement thrombotique. Lettre aux professionnels de santé” May 2019: 3 pages.  2- “Thrombophilia. Testing rarely useful after a venous thromboembolic event” Prescrire Int 2017; 26 (182): 129.  3- “Oral anticoagulants in atrial fibrillation. Warfarin or apixaban, depending on the clinical situation” Prescrire Int 2019; 28 (205): 159-160.

● Rivaroxaban is less effective than warfarin in preventing thrombosis in patients with anti- phospholipid antibodies, a risk factor for thrombo- embolic events. In mid-2019, the French drug regulatory agency, ANSM, advised health professionals not to use direct oral anticoagulants in patients with a history of thrombosis and known to have antiphospho- lipid antibodies ( a ). Switching to a vitamin K antag- onist such as warfarin should be considered for such patients who are already taking a direct oral anticoagulant (1). The presence of antiphospholipid antibodies characterises the most common acquired thrombo- philia. These antibodies are detected in 1% to 6% of the population. Persons with antiphospholipid antibodies have a 3- to 15-fold higher risk of venous thrombosis than the general population. They are also at increased risk of arterial thrombosis, includ- ing ischaemic stroke (2). More thromboembolic events with rivaroxaban. This warning is based on the results of a non- blinded clinical trial of rivaroxaban versus warfarin in 120 patients in whom a history of thrombosis led to the detection of antiphospholipid antibodies.The trial was stopped prematurely due to a higher inci- dence of thromboembolic events in the rivaroxaban group: 12% (4 cases of ischaemic stroke and 3 cases of myocardial infarction), versus no cases in the warfarin group (1). Given that other direct oral anticoagulants could pose the same risk, and in the absence of data to the contrary as of mid-2019, the warning also applies to dabigatran, apixaban and edoxaban (1).

Prescrire Int • February 2020

Sign up online at english.prescrire.org to receive the free, monthly "Independent Drug & Healthcare Newsletter"

From the non-profit journal Prescrire International , to keep you informed on the topics covered in every issue

P age 10 • P rescrire I nternational S pecial E dition 2020

ADVERSE EFFECTS

Antiepileptics and pregnancy: potential long-term effects in children


ABSTRACT

● In summary, as of mid-2019, no antiepileptic drugs are risk-free for the unborn child. Lamo- trigine is the antiepileptic on which the most data are available. Lamotrigine appears to be associat- ed with a lower risk of learning disabilities than the other antiepileptics. The long-term effects of pregabalin , carbamazepine , topiramate , leveti- racetam and gabapentin remain highly uncertain. Valproic acid and its derivatives are teratogenic and provoke serious disorders of neuropsycho­ logical and psychomotor development. Rev Prescrire 2019; 39 (430): 587-595 A ntiepileptic drugs are used in various forms of epilepsy; some are also used in the treat- ment of certain psychiatric disorders or pain syndromes. Many women who could become pregnant are exposed to these drugs. Yet most of the antiepileptic drugs taken by pregnant women expose the unborn child to the risk of malformations, fetotoxic effects and neonatal disorders (see the insets “Short-term effects of in utero exposure to antiepileptics” and “Valproic acid: the consequenc- es of in utero exposure”, Prescrire Int n° 211, pp. 16-19). Antiepileptics also have psychotropic prop- erties and could therefore interfere with the psy- chomotor and behavioural development of children exposed in utero. For example, valproic acid can delay psychomotor development and cause be- havioural disorders in the long term.These serious adverse effects must be weighed against the harm- ful consequences of untreated maternal epilepsy on the pregnancy and its outcome: preeclampsia, bleeding, fetal growth restriction, preterm birth, miscarriage, and maternal or fetal death. Prescrire reviewed the risk of malformations and long-term effects in children exposed in utero to valproic acid , carbamazepine , lamotrigine , phenytoin and topiramate in 2009, and has since reported on additional findings concerning these risks. What are the main data available in 2019 on the long-term consequences for children exposed in utero to these antiepileptic drugs? ▶ Excerpt from Rev Prescrire August 2019 Volume 39 N° 430 • Page 587-595 Full review (8 pages) available at english.prescrire.org for subscribers

● What are the main data available in mid- 2019 that compare the long-term effects of the most widely used antieleptics on children exposed to these drugs in utero? ● Two systematic reviews with meta-analysis and four studies published since 2017, including a large French cohort study, provide information about the long-term risk of disorders of neurological and behavioural development in children exposed in utero to monotherapy with various antiepileptic drugs. The comparator groups were children born to women with the same disease but who received either no treatment, or placebo or other antiepilep- tics; or children born to women without epilepsy. ● With lamotrigine , an approximately 1.6-fold increased risk of early developmental disorders was demonstrated in about 3000 children includ- ed in the French cohort study who had been exposed to this antiepileptic in utero and followed up until the age of 5 years.A meta-analysis of data from almost 300 pregnancies showed an approxi­ mately 9-fold increased risk of autism and dys­ praxia with lamotrigine ; other studies in fewer patients yielded conflicting results. ● Very limited data are available on pregabalin , carbamazepine, topiramate , gabapentin and leve- tiracetam . Pregabalin may increase the risk of intel- lectual disability, mental disorders and behavioural disorders. A 2-fold increased risk of disorders of early mental development was observed with car- bamazepine . ● Limited data on topiramate in 2014 showed a long-term impact on motor development and cog- nitive and visual function, sometimes with mal- adaptive behaviour. The 2018 French study also showed a risk of learning disabilities in the 500 chil- dren exposed in utero to topiramate . ● With gabapentin , there are concerns over a potential risk of early mental, behavioural and developmental disorders. ● Data are available on too few pregnancies exposed to levetiracetam to determine its long- term consequences.

Prescrire Int • January 2020

P rescrire I nternational S pecial E dition 2020 • P age 11