Drugs to Avoid in 2016

Outlook

Translated from Rev Prescrire 2016; 36 (388): 138-146 Towards better patient care: drugs to avoid in 2016

Abstract

To help healthcare professionals and• patients choose high-quality treatments that minimize the risk of adverse effects, in early 2016 we updated our list of drugs to avoid.

Prescrire’s assessments of the harm-benefit• balance of new drugs and indications are based on a rigor- ous procedure that includes a system- atic and reproducible literature search, identification of patient-relevant out- comes, prioritisation of the supporting data based on the strength of evi- dence, comparison with standard treatments, and an analysis of both known and potential adverse effects.

This 2016 review of medications examined• by Prescrire over a six-year period, from 2010 to 2015, identified 74 drugs that are more harmful than beneficial in all the indications for which they have been authorised in France. A reliable, rigorous and The editorial staff comprise a broad independent methodology range of health professionals working In most cases, when drug therapy in various sectors and free of conflicts is• really necessary, other drugs with a What data sources and methodology of interest. We also call on an extensive better harm-benefit balance are avail- do we use to assess the harm-benefit network of external reviewers (special- able. balance of a given drug? ists, methodologists, and practitioners The following review concerns drugs representative of our readership), and Even in serious situations, when no and indications on which we pub- each article undergoes multiple qual­ity effective• treatment exists, there is no lished detailed analyses in our French controls and cross checking at each justification for prescribing a drug with edition over a six-year period, from step of the editorial process (see About no proven efficacy that provokes 2010 to 2015. Some drugs and indica- Prescrire > How we work at english.­ severe adverse effects. It may be tions were examined for the first time, prescrire.org). Our editorial process is acceptable to test these drugs in clin- while others were re-evaluated as new a collective one, as symbolised by the ical trials, but patients must be data on efficacy or adverse effects “©Prescrire” signature. informed of the uncertainty over their became available. Most importantly, Prescrire is fiercely harm-benefit balance, and the trial The overriding goal of Association independent. Our work is funded sole- design must be relevant. Tailored sup- Mieux Prescrire, the not-for-profit ly and entirely by our subscribers. No portive care is the best option when organisation that publishes the jour- company, professional organisation, there are no available treatments cap­ nals la revue Prescrire and Prescrire Inter­ insurance system, government agency able of improving prognosis or quality national, is “to work in total independence or health authority has any financial of life, beyond the placebo effect. to promote quality healthcare, first and influence whatsoever over the con- Rev Prescrire 2016; 36 (388): 138-146. foremost in the interest of patients” (Article tents of our publications. 1 of the bylaws). All our publications his is Prescrire’s fourth consecutive are intended to provide health profes- Comparison with standard treat- annual review of “drugs to avoid” sionals (and their patients) with the ments. The harm-benefit balance of a T(1,2). The listed drugs are clearly clear, independent, reliable and up-to- given drug has to be continually more dangerous than beneficial and date information they need, free of re-evaluated as new data on efficacy or should therefore not be used in any conflicts of interest and commercial adverse effects become available. Like- circumstance. The aim is to help health pressures. wise, treatment options evolve as new professionals to choose safe, effective Prescrire is structured in such a way drugs arrive on the market. treatments and thereby avoid harming as to guarantee the quality of the infor- Not all drugs are equal: some offer their patients. mation provided to our subscribers. a therapeutic advantage, while

Prescrire International April 2016/Volume 25 N° 170 • Page 105 Outlook Towards better patient care: drugs to avoid in 2016

­others are more harmful than ben- Serious conditions with no effec- used and then in alphabetical order of eficial and should not be used (3). tive treatment: patients should be their international nonproprietary All Prescrire’s assessments of new informed of the consequences of names (INNs). drugs and indications are based on a interventions. When faced with a These 74 drugs comprise: systematic and reproducible literature serious condition for which there is no – Active substances with adverse search. The resulting data are then effective treatment, some patients opt effects that are disproportionate to the analysed collectively by our editorial to forgo treatment while others are benefits they provide; staff, using an established procedure: willing to try any drug that might – Older drugs that have been super- – Efficacy data are prioritised: most bring them even temporary relief, seded by new drugs with a better weight is given to studies providing despite a risk of serious adverse effects. harm-benefit balance; robust supporting evidence, i.e. When the short-term prognosis is – Recent drugs that have a less favour- well-conducted, double-blind, ran- poor, some health professionals may able harm-benefit balance than exist- domised controlled trials; propose last-chance treatments with- ing options; – The new drug is compared with a out properly informing the patient of – Drugs that have no proven efficacy carefully chosen standard treatment the harms, either intentionally or (beyond the placebo effect) but that (not necessarily a drug); unwittingly. carry a risk of serious adverse effects. – The accent is placed on those clinical Yet patients in this situation must The main reasons why these drugs endpoints most relevant to the patients not be treated as guinea pigs. It is very are considered to have an unfavour- concerned. This means that we often useful to enrol patients into clinical able harm-benefit balance are ignore surrogate endpoints such as research provided they are informed explained in each case. When avail- simple laboratory markers that have of the harms and the uncertain nature able, better options are briefly men- not been shown to correlate with a of the possible benefits, and that the tioned, as are situations (serious or favourable clinical outcome (4,5). results are published in order to non-serious) in which there is no suit- advance medical knowledge. able treatment. Careful analysis of adverse But patients must be aware that they The differences between this year’s effects. Adverse effects can be more are free to refuse to participate in clin- and last year’s lists are detailed in the difficult to analyse, as they are often ical trials or to receive last-chance inset on p. 107. less thoroughly documented than effi- treatments with an uncertain cacy, and this discrepancy must be harm-benefit balance. They must also taken into account. be reassured that, if they do refuse, Antibiotics The adverse effect profile of each they will not be abandoned but con- drug is assessed by examining data from tinue to receive the best available care. • Moxifloxacin is no more effective clinical trials and animal pharmaco­ Even though they are not aimed at than other fluoroquinolone antibiotics toxicology studies, and any pharmaco- modifying the outcome of the under- but can cause toxic epidermal necroly- logical affiliation. lying disease, supportive care and sis and fulminant hepatitis and has Marketing authorisation of a new symptomatic treatment are key ele- also been linked to an increased risk of drug does not signify that its harm-­ ments of patient care. cardiac disorders (Prescrire Int n° 62, benefit balance has been fully docu- By their very nature, clinical trials 103; Rev Prescrire n° 371). Another mented. Indeed, rare but serious involve a high degree of uncertainty. fluoroquinolone­ such as ciprofloxacin or adverse effects may only emerge after In contrast, drugs used for routine care ofloxacin is a better option. several years of routine use (3). must have an acceptable harm-benefit • Telithromycin has no advantages balance. Marketing authorisation over other macrolide antibiotics but Empirical data and personal should only be granted on the basis of carries an increased risk of QT interval experience: risk of bias. Empirical proven efficacy relative to a standard prolongation, hepatitis, visual distur- assessment of a drug’s harm-benefit treatment, and an acceptable adverse bances and syncope (Prescrire Int n° 84, balance based on individual experi- effect profile: in general, little extra 88, 94, 106, 154). Another macrolide ence can help to guide further research information on efficacy is collected such as spiramycin is a better option. but is subject to major bias and rep- once marketing authorisation has been resents only weak evidence (3,4). For granted (3). example, it can be difficult to attribute Cardiology a specific outcome to a particular drug, as other factors must be taken into 74 drugs more dangerous than • Aliskiren, an antihypertensive renin account, including the natural history beneficial inhibitor, has not been shown to pre- of the disease, the placebo effect, the vent cardiovascular events. On the effect of another treatment the patient This review lists drugs that have an contrary, a trial in diabetic patients may not have mentioned, or a change unfavourable harm-benefit balance in showed that aliskiren was associated in lifestyle or diet. Similarly, a doctor all their authorised indications, in other with an excess of cardiovascular events who sees an improvement in certain words drugs that should be removed and renal failure (Prescrire Int n° 106, patients may be unaware that many from the market on account of their 129, 166). It is better to choose one of other patients have been harmed by toxicity. Drugs with an unfavourable the many tried-and-tested antihyper- the same treatment (3). harm-benefit balance in certain situa- tensive drugs such as a thiazide diuret- The best way to overcome this tions but not in others have not been ic or an angiotensin converting ­subjective bias due to non-comparative­ included. enzyme (ACE) inhibitor. evaluation of a few patients is to Between 2010 and 2015, we identi- • Bezafibrate, ciprofibrate and fenofi- ­prioritise well-conducted clinical fied 74 drugs marketed in France that brate are cholesterol-lowering drugs ­studies, particularly double-blind, are more dangerous than beneficial. with no proven efficacy in the preven- ­randomised trials versus a standard They are listed below, based first on tion of cardiovascular events (beyond ­treatment (3,4). the therapeutic area in which they are the placebo effect), yet they all have

Page 106 • Prescrire International April 2016/Volume 25 N° 170 numerous adverse effects, including cutaneous, haematological and renal Notable changes in the 2016 update: disorders (Prescrire Int n° 85, 117). When citalopram, escitalopram, diclofenac added a fibrate is justified,gemfibrozil is the only one that has been shown to pre- to the list of drugs to avoid vent cardiovascular complications of hypercholesterolaemia, although renal Three of the drugs that have featured in than many other antidepressants over function and serum creatine phosphoki- our list of drugs to avoid since the first which it has no proven advantages (Rev nase levels must be closely monitored. version, published in 2013, were with- Prescrire n° 386 and Prescrire Int n° 170). • Ivabradine, an inhibitor of the car- drawn from the French market in 2015 by Omalizumab, which is authorised for use diac If current, can cause visual distur- the pharma­ceutical companies concerned: in asthma and chronic spontaneous urti- bances, cardiovascular disorders asenapine for manic episodes; iron dextran caria, is no more effective than a cortico- (including myocardial infarction), for anaemia; and floctafenine for moderate steroid. In addition to its immunosuppres- potentially severe bradycardia and pain. sant effect, this monoclonal antibody other cardiac arrhythmias. It has no causes hypersensitivity reactions and advantages in angina or heart failure Pirfenidone: not listed in 2016, but cardiac disorders (Prescrire Int n° 115, (Prescrire Int n° 88, 110, 118, 155, many uncertainties. All the drugs listed 161). 165). Established treatments shown to in our 2015 review are also included this be effective in angina include year, with the exception of pirfenidone, Additions: drugs that are more harm- beta-blockers and the calcium channel whose harm-benefit balance in idiopathic ful than similar options. An analysis of blockers amlodipine and verapamil. pulmonary fibrosis has become more the cardiac adverse effects of antidepres- There are also better options for heart uncertain in light of new clinical data. Its sants revealed that the “selective” sero- failure: one is to refrain from adding clinical evaluation includes some favour- tonin reuptake inhibitors (SSRIs) citalo- another drug to an optimised treat- able data but still does not show whether pram and escitalopram are no more ment regimen; another is to use a or not pirfenidone reduces mortality, even effective than other SSRIs but cause more beta-blocker with a proven impact on after one year. It is not clear whether the cardiac disorders, including dose-­ mortality. uncertain benefit of this treatment out- dependent prolongation of the QT interval • Nicorandil, a vasodilator with solely weighs its harms, which markedly reduce and torsades de pointes (Rev Prescrire symptomatic efficacy in the prevention the quality of life of patients whose life n° 386). of effort angina, can cause severe expectancy is short, but this does not jus- Analysis of the cardiovascular adverse mucocutaneous ulceration (Prescrire Int tify its continued inclusion in our list of effects of nonsteroidal anti-inflammatory n° 81, 95, 110, 132). A nitrate is a drugs to avoid (Rev Prescrire n° 384). drugs (NSAIDs) revealed that diclofenac better option to prevent angina attacks. causes more cardiovascular adverse • Olmesartan, an angiotensin II recep- Confirmation: thiocolchicoside, ven- effects, including myocardial infarction, tor blocker (ARB or sartan) that is no lafaxine, omalizumab. In 2015, we re-­ heart failure and cardiovascular deaths more effective than other ARBs in examined certain aspects of the harm-­ than other NSAIDs, such as ibuprofen (up hypertension, can cause sprue-like benefit balance of several drugs from our to a maximum dose of 1200 mg per day) enteropathy with chronic diarrhoea list of drugs to avoid. Our re-evaluation of or naproxen, but is no more effective. In the (potentially severe) and weight loss, thiocolchicoside, a drug with a similar absence of evidence to the contrary, and, possibly, an increased risk of car- chemical structure to colchicine, confirmed ­aceclofenac was considered to expose diovascular mortality (Prescrire Int its place on the list. Thiocolchicoside has patients to similar risks to diclofenac due n° 148). It is better to choose another a variety of serious hepatic, pancreatic, to their chemical affiliation, and should of the many available ARBs, such as muscular, haematological and neurological therefore also be avoided (Prescrire Int losartan or valsartan, which do not adverse effects, yet has not been shown n° 167; Rev Prescrire n° 374). appear to have these adverse effects to be more effective than placebo in mus- The efficacy ofdefibrotide , an antithrom- • Trimetazidine, a drug with uncer- cle pain ­(Prescrire Int n° 168). botic authorised in severe hepatic veno-­ tain properties, is used in angina Re-analysis also confirmedvenlafaxine occlusive disease following haemopoietic despite its only modest symptomatic as an antidepressant to be avoided. This stem cell transplantation, is too uncertain efficacy (shown mainly in stress antidepressant with serotonergic and nor- when balanced against its serious adverse tests), yet it can cause parkinsonism, adrenergic activity causes more cardiovas- effects, in particular haemorrhages hallucinations and thrombocytopenia cular adverse effects, and is more likely to ­(Prescrire Int n° 164). (Prescrire Int n° 84, 100, 106). It is result in death in the event of overdose, ©Prescrire better to choose better-known treat- ments for angina, such as certain beta-blockers or the calcium-channel blockers amlodipine and verapamil. co-administration of drugs that inhibit inadvertent injection into an artery (Rev this isoenzyme (Rev Prescrire n° 337). A Prescrire n° 327). Injectable dexchlor­ non-sedating antihistamine without pheniramine, which does not appear to Dermatology - Allergy antimuscarinic activity, such as lorata­ carry these risks, is a better option. dine or cetirizine, is a better option in • Topical tacrolimus, an immunosup- • Mequitazine, a sedating antihista- this situation. pressant used in atopic eczema, can mine with antimuscarinic properties, • Omalizumab in chronic spontaneous cause skin cancer and lymphoma, yet its used in allergies, has only modest effi- urticaria (see the Pulmonology - ENT efficacy is barely different from that of cacy but carries a higher risk than other section on p. 111) (Prescrire Int n° 161). topical corticosteroids (Prescrire Int n° antihistamines of cardiac arrhythmias • Injectable promethazine, an antihis- 101, 110, 131; Rev Prescrire n° 367). Judi- due to QT prolongation in patients who tamine used to treat severe urticaria, cious use of a topical corticosteroid to are slow cytochrome isoenzyme P450 can cause thrombosis, skin necrosis and treat flare-ups is a better option in this CYP2D6 metabolisers, and during gangrene following extravasation or situation.

Prescrire International April 2016/Volume 25 N° 170 • Page 107 Outlook Towards better patient care: drugs to avoid in 2016

Diabetes - Nutrition it is better to use metoclopramide, very • Memantine, an NMDA glutamate carefully, at the lowest possible dose receptor antagonist, can cause neuro- • Gliptins inhibit dipeptidylpepti- and for the shortest possible ­period. psychiatric disorders such as hallucin­ dase-4 (DPP-4), an enzyme that catab- • Prucalopride, a drug chemically ations, confusion, dizziness, headache olises the gut hormones (incretins) ­related to neuroleptics, is authorised for (creating a risk of violent behaviour) that stimulate postprandial insulin chronic constipation but shows only * and seizures (Rev Prescrire n° 359, 362, secretion, but these drugs have no modest efficacy, in about one in six 374). proven efficacy against the complica- patients. Its adverse effect profile is tions of diabetes (cardiovascular poorly documented, particularly with Multiple sclerosis. The standard events, renal failure, neurological dis- respect to cardiovascular disorders “disease-modifying” treatment for orders, etc.). This is the case of lina- ­(palpitations, ischaemic cardiovascular multiple sclerosis is interferon beta, gliptin, saxagliptin, sitagliptin and events, possible QT prolongation) and despite its limitations and many vildagliptin, whether used alone or in teratogenicity (Prescrire Int n° 116, 137). adverse effects. The harm-benefit bal- combination with metformin. These There is no justification for exposing ance of other such treatments is no four drugs have an unfavourable patients with simple constipation to better and sometimes clearly unfa- adverse effect profile that includes such risks. If dietary meas­ures are vourable. This is particularly the case severe hypersensitivity reactions (ana- ­ineffective, then bulk-forming laxatives, of two immunosuppressants, which phylaxis, Stevens-Johnson syndrome), osmotic laxatives or, very occasionally, have disproportionate adverse effects infections (urinary tract and upper other laxatives (lubricants, stimulants, and should be avoided. respiratory tract infections), pancrea­ or rectal preparations), used carefully • Natalizumab, a monoclonal anti- titis, bullous pemphigoid and intestinal and patiently, are safer than prucalopride. body, can lead to life-threatening obstruction (Prescrire Int n° 121, 135, opportunistic infections, including 138, 158, 167; Rev Prescrire n° 365, progressive multifocal leukoencephal­ 366). A proven treatment such as met­ Gynaecology - Endocrinology opathy (in about 2 per 1000 patients), formin, or glibenclamide or insulin if met­ potentially severe hypersensitivity formin is insufficiently effective, or • Tibolone, a synthetic steroid hor- reactions and liver damage (Rev targeting a higher HbA1c, are more mone used for postmenopausal ­Prescrire n° 330, 333, 374). reasonable choices. replacement therapy, has androgenic, • Teriflunomide has potentially • Orlistat has only modest and tran- oestrogenic and progestogenic proper- life-threatening adverse effects, includ- sient efficacy in terms of weight loss ties and carries a risk of cardiovascular ing liver damage, leukopenia and (about 3.5 kg more than placebo after disorders, breast cancer and ovarian infections. There is also a risk of 12 to 24 months). There is no evidence cancer (Prescrire Int n° 83, 11, 137). peripheral neuropathy (Rev Prescrire of long-term efficacy. Gastrointestinal When hormone therapy is chosen n° 373). disorders are very frequent, while despite the inherent risks, the most other adverse effects include liver reasonable option is an oestrogen-­ Miscellaneous. A number of drugs damage, hyperoxaluria and bone frac- progestogen combination, used at the used in migraine and Parkinson’s dis- tures in adolescents. Orlistat alters the lowest possible dose and for the short- ease should also be avoided. gastrointestinal absorption of many est possible period. • Flunarizine and oxetorone, two nutrients (fat-soluble vitamins A, D, E neuroleptics used to prevent migraine and K), leading to a risk of deficiency, attacks, have at best only modest effi- and also reduces the efficacy of some Neurology cacy (flunarizine prevents about one drugs (thyroid hormones, some anti­ attack every two months) but can epileptics). Oral contraceptive efficacy Alzheimer’s disease. The drugs cause extrapyramidal disorders, car­ can be reduced if orlistat provokes available for Alzheimer’s disease in diac disorders and weight gain severe diarrhoea (Prescrire Int n° 57, early 2016 have only minimal and (Prescrire­ Int n° 137). It is better 71, 107, 110; Rev Prescrire n° 374). transient efficacy. They are also diffi- to choose another drug such as There are currently no drugs capable cult to use because of their dispropor- propranolol­ . of inducing permanent weight loss. It tionate adverse effects and many inter- • Tolcapone, an antiparkinsonian drug, is better to focus on dietary changes actions with other drugs. None of the can cause life-threatening liver damage and physical activity. available drugs has been shown to slow (Rev Prescrire n° 330). When other progression toward dependence, yet all treatment options have been exhaust- carry a risk of life-threatening adverse ed, entacapone is a better option. Gastroenterology effects and severe drug interactions (Prescrire Int n° 128; Rev Prescrire n° 363, • Domperidone and droperidol, two 364). It is better to focus on reorganis- Oncology - Haematology neuroleptics, can cause ventricular ing the patient’s daily life, keeping him arrhythmias and sudden death, which or her active and providing support • Catumaxomab, used in malignant are disproportionate to the symptoms and help for caregivers. ascites, has serious adverse effects (pos- and their weak efficacy against nausea • Donepezil, galantamine and rivastig- sibly fatal) in more than three-quarters and vomiting, and, for domperidone, mine, three cholinesterase inhibitors, of patients (Prescrire Int n° 109). Para- against gastro-oesophageal reflux can cause gastrointestinal disorders centesis is a better option, repeated as (Prescrire Int n° 129, 144; Rev Prescrire (including severe vomiting), neuro- necessary to control symptoms. n° 365, 371). Other drugs such as ant- psychiatric disorders, cardiac disorders • Defibrotide, an antithrombotic acids and omeprazole have a much better­ (including bradycardia, malaise and approved to treat severe hepatic harm-­benefit balance in gastro-­ syncope) and cardiac conduction dis- veno-occlusive disease following oesophageal reflux disease. In the rare orders; galantamine can cause serious haemopoietic stem cell transplanta- situations in which treatment with an * skin disorders (Prescrire Int n° 162, 166; tion, had no more impact on mortal- antiemetic neuroleptic appears justified, Rev Prescrire n° 337, 340, 344, 349). ity or complete disease remission

Page 108 • Prescrire International April 2016/Volume 25 N° 170 * Correction made after initial publication than symptomatic treatment in an Pain - Rheumatology vertebral fractures. Yet its adverse unblinded trial, but provokes some- effects include neuropsychiatric disor- times fatal haemorrhages (Prescrire • Analgesics. Many nonsteroidal ders, cardiovascular disorders (includ- Int n° 164). It is better to focus on anti-inflammatory drugs (NSAIDs) ing venous thrombosis and pulmonary preventive measures and symptom- should be avoided, especially since embolism, myocardial infarction and atic treatments. alternatives with a better harm-benefit cardiovascular death), and hypersensi- • Panitumumab does not prolong balance are available. Paracetamol is the tivity reactions including toxic epider- survival in metastatic colorectal can- first-choice analgesic: it is effective for mal necrolysis and DRESS syndrome cer, yet about 90% of patients experi­ moderate pain and poses little danger (Drug Reaction with Eosinophilia and ence adverse effects, which include when taken at the appropriate dosage. Systemic Symptoms) (Prescrire Int severe skin damage (sometimes Certain NSAIDs, such as ibuprofen and n° 117, 125, 139, 142, 156). resulting in fatal infections), gastro­ naproxen, used at the lowest effective intestinal and ocular disorders, inter- dose and for the shortest possible peri- Osteoarthritis. Drugs authorised as stitial lung disease and hypersensitiv- od, are an alternative. disease-modifying osteoarthritis drugs ity reactions (Prescrire Int n° 138). It is • Cox-2 inhibitors (coxibs) such as should be avoided because they have unreasonable to add panitumumab to celecoxib, etoricoxib and parecoxib significant adverse effects but no prov- tried-and-tested chemotherapy regi- have been linked to an excess of car- en efficacy beyond the placebo effect. mens such as those based on fluo­ diovascular events (including myocar- A better option in this situation is rouracil, alone or combined with other dial infarction and thrombosis) and paracetamol as the first-choice treat- cytotoxic drugs. skin reactions by comparison with ment for pain, when taken at the • Trabectedin showed no tangible effi- other, equally effective NSAIDs appropriate dosage. Carefully chosen cacy in comparative trials in ovarian ­(Prescrire Int n° 167; Rev Prescrire and closely monitored nonsteroidal cancer and soft-tissue sarcomas but n° 344, 361, 374). anti-inflammatory drug therapy is has very frequent and severe gastro- • Diclofenac and aceclofenac cause sometimes an acceptable option. intestinal, haematological, hepatic and more cardiovascular adverse effects • Diacerein causes gastrointestinal dis- muscular adverse effects (Prescrire Int (including myocardial infarction and orders (including gastrointestinal n° 102, 120; Rev Prescrire n° 360). It is heart failure) and cardiovascular bleeding and melanosis coli), angioede- unreasonable to add trabectedin to deaths than other, equally effective ma and hepatitis (Rev Prescrire ­platinum-based chemotherapy for NSAIDs (Prescrire Int n° 167; Rev n° 282, 321; Prescrire Int n° 159). ovarian cancer. When chemotherapy ­Prescrire n° 362, 374). • Glucosamine causes allergic reac- is ineffective in patients with soft-­ • Ketoprofen gel causes more photo- tions (angioedema, acute interstitial tissue sarcomas, it is best to focus on sensitivity reactions (eczema, bullous nephritis) and hepatitis (Prescrire Int appropriate supportive care. rash) than other, equally effective top- n° 84, 137; Rev Prescrire n° 380). • Vandetanib has no proven impact ical NSAIDs (Prescrire Int n° 109, 137). on survival in patients with metastatic • Piroxicam, when used systemically, Miscellaneous. A number of other or inoperable medullary thyroid can- is associated with an increased risk of drugs used primarily in rheumatology cer. Too many patients were lost to gastrointestinal and cutaneous disor- should be avoided. follow-up in placebo-controlled trials ders (including toxic epidermal • Muscle relaxants with no proven to show an increase in progression-free necrolysis (Lyell’s syndrome)) but is efficacy beyond the placebo effect: survival. Serious adverse effects (diar- not more effective than other NSAIDs methocarbamol has many adverse rhoea, pneumonia, hypertension) (Rev Prescrire n° 321). effects, including gastrointestinal and occur in about one-third of patients. cutaneous disorders (angioedema); There is also a risk of interstitial lung Osteoporosis. Several drugs thiocolchicoside, which is related to disease, torsades de pointes and sud- ­authorised for osteoporosis should be colchicine, causes diarrhoea, stomach den death (Prescrire Int n° 131). Here avoided because their efficacy is at best pain, photodermatosis and possibly too it is best to focus on tailored sup- modest and they have potentially seri- convulsions; it is also genotoxic and portive care. ous adverse effects. When non-drug teratogenic (Rev Prescrire n° 282, 321, • Vinflunine has uncertain efficacy in measures plus calcium and vitamin D 313, 367; Prescrire Int 168). There is no advanced and metastatic bladder can- supplementation prove inadequate, justification for exposing patients with cer. A clinical trial provided weak alendronic acid or an alternative, ralox­ simple muscle pain to these adverse evidence that vinflunine increases ifene, have a better harm-benefit bal- effects. An effective analgesic such as median survival by two months at ance than other options, despite the paracetamol is a better option, when best compared with palliative care. significant limitations of both drugs. taken at the appropriate dosage. There is a high risk of haematological • Denosumab 60 mg in osteoporosis • Pegloticase, a recombinant urate oxi- adverse effects (including aplastic has very modest efficacy in the pre- dase used in severe gout, has modest anaemia), and a risk of serious infec- vention of osteoporotic fractures and short-term symptomatic efficacy and tions and cardiovascular disorders no efficacy for “bone loss” during pros- disproportionate adverse effects, (torsades de pointes, myocardial tate cancer, but carries a disproportion- including severe reactions during infu- infarction, ischaemic heart disease), ate risk of adverse effects, including sion (despite premedication), anaphy- sometimes resulting in death (­ Prescrire back pain, musculoskeletal pain, and laxis, severe skin infections and, pos- Int n° 112). When platinum-based serious infections (including endocar- sibly, severe cardiac disorders (Rev chemotherapy is ineffective, it is best ditis) due to the immunosuppressive Prescrire n° 365). When treatment to focus on tailored supportive care. effects of this monoclonal antibody (Prescrire Int n° 117, 130, 168). There is no known satisfactory­ drug treat- a- A 120-mg strength denosumab product is notably ment for “bone loss” (a). authorised for bone metastases from solid tumours. In this setting, denosumab offers no tangible clinical advan­ • Strontium ranelate has only modest tage, but its harms do not clearly outweigh its benefits efficacy in the prevention of recurrent (Prescrire Int n° 130).

Prescrire International April 2016/Volume 25 N° 170 • Page 109 Outlook Towards better patient care: drugs to avoid in 2016

with the first-choice drug allopurinol (SSRIs) but also carries a risk of cardiac Other psychotropic drugs. Some and the alternative probenecid is inad- disorders (hypertension, tachycardia, other psychotropic drugs have unac- equate or risky, it is better to manage arrhythmias, etc.) due to its noradren- ceptable adverse effects. attacks with symptomatic treatments, ergic activity. Duloxetine can also cause • Dapoxetine, a “selective” SRI, is used pending a better solution. hepatitis and severe cutaneous hyper- for premature ejaculation with sexual • Quinine, used to treat cramps, can sensitivity reactions such as Stevens-­ dissatisfaction. Its adverse effects are have life-threatening adverse effects Johnson syndrome (Prescrire Int n° 85, disproportionate to its very modest including anaphylactic reactions, 100, 111, 142; Rev Prescrire n° 384). efficacy and include aggressive out- ­haematological disorders (including • Citalopram and escitalopram are bursts, serotonin syndrome and syn- thrombocytopenia, haemolytic anae- SSRI antidepressants that increase the cope (Prescrire Int n° 105; Rev ­Prescrire mia, agranulocytosis, and pancytope- incidence of QT prolongation and tor- n° 355). A psychological and nia) and cardiac arrhythmias. These sades de pointes compared with other behavioural approach is a better option adverse effects are disproportionate in SSRI antidepressants (Rev Prescrire in this situation. view of its poor efficacy Rev( Prescrire n° 369, 386). • Etifoxine, a drug poorly evaluated in n° 337, 344). There are no drugs with • Milnacipran and venlafaxine, two anxiety, can cause hepatitis and severe a favourable harm-benefit balance for non-tricyclic, non-SSRI, non-­ hypersensitivity reactions (including patients with cramps. Stretching is monoamine oxidase inhibitor (MAOI) DRESS syndrome, Stevens-Johnson sometimes beneficial Rev( Prescrire antidepressants, have both serotoner- syndrome and toxic epidermal n° 363). gic and noradrenergic activity. Not necrolysis) (Prescrire Int n° 136; Rev • Colchimax° (colchicine + opium only do they have the adverse effects Prescrire n° 376). When an anxiolytic powder + tiemonium) should be avoid- of SSRI antidepressants, they also drug is justified, it is better to choose a ed in gout attacks because the action cause cardiac disorders (hypertension, benzodiazepine, for the shortest pos­ of powdered opium and tiemonium can tachycardia, arrhythmias, QT prolon- sible period. mask the onset of diarrhoea, which is gation) due to their noradrenergic an early sign of potentially fatal colchi­ activity. In addition, venlafaxine over- Smoking cessation. Some drugs cine overdose (Prescrire Int n° 147). A doses are associated with a high risk of authorised to assist with smoking cessa- nonsteroidal anti-inflammatory drug, cardiac arrest (Rev Prescrire n° 338, 343, tion are no more effective than nico­tine or colchicine alone, are better options in 386 and Prescrire Int n° 170). and have more adverse effects. When a this situation. • Tianeptine, a drug with no proven drug is needed to help with smoking • The dexamethasone + salicylamide efficacy, can cause hepatitis, life-­ cessation, nicotine is a better choice. + hydroxyethyl salicylate combination threatening skin reactions (including • Bupropion, an amphetamine, can (Rev Prescrire n° 345) and the prednis- bullous rash), abuse and addiction cause neuropsychiatric disorders olone + dipropylene glycol salicylate (Prescrire Int n° 127, 132). (including aggressiveness, depression combination (Rev Prescrire n° 338), when applied to the skin, expose patients to the adverse effects of cortico­steroids and to salicylate hyper- sensitivity reactions. Other drugs such as oral paracetamol (at the appropriate dosage) and topical ibuprofen have a better harm-benefit balance in patients with painful sprains or tendinopathy, in conjunction with non-drug meas­ ures (rest, ice, splints).

Psychiatry - Addiction

Antidepressants. Several drugs authorised for depression carry a greater risk of severe adverse effects but are no more effective than alter- native treatments. In general, anti­ depressants have only modest efficacy and often take some time to work. It is better to choose an antidepressant with an adequately documented adverse effect profile. • Agomelatine has no proven efficacy beyond the placebo effect, but can cause hepatitis and pancreatitis, suicide and aggression, as well as serious skin disorders including Stevens-Johnson syndrome (Prescrire Int n° 136, 137). • Duloxetine, a serotonin and norepi- nephrine reuptake inhibitor, not only has the adverse effects of the so-called “selective” serotonin reuptake inhibitors

Page 110 • Prescrire International April 2016/Volume 25 N° 170 and suicidal ideation), potentially dose is a better option in both of these what justification is there for exposing severe allergic reactions (including situations. patients to drugs with severe adverse angioedema and Stevens-­Johnson • Pholcodine, an opioid used as an effects but no proven impact (beyond syndrome), addiction, and congenital antitussive, can cause sensitisation to the placebo effect) on patient-relevant heart defects in children exposed to neuromuscular blocking agents (Rev clinical outcomes? the drug in utero (Prescrire Int n° 131; Prescrire n° 349). This serious adverse It is necessary but not sufficient for Rev Prescrire n° 377). effect is not known to occur with other health professionals to remove these • Varenicline can cause depression, opioids. Cough is a minor ailment that drugs from their list of useful treat- suicide, severe skin rash (including does not warrant taking such risks. ments: regulators and health authori- Stevens-Johnson syndrome) and car- When drug therapy is required for ties must also take concrete steps to diac disorders (angina, myocardial cough, it is better to choose dextro­ protect patients and promote the use infarction, atrial fibrillation) Prescrire( methorphan, despite its limitations (Rev of treatments that have an acceptable Int n° 124, 131; Rev Prescrire n° 377). Prescrire n° 358). harm-benefit balance. • Tixocortol (sometimes combined The drugs listed above are more dan- with chlorhexidine), a corticosteroid gerous than beneficial. There is no Pulmonology - ENT authorised for sore throat, can cause valid reason for them to remain on the allergic reactions such as facial muco- market. • Oral and nasal vasoconstrictive cutaneous oedema, glossitis and even Review produced collectively by the decongestants (ephedrine, naphazo- angioedema (Rev Prescrire n° 320). Editorial Staff: no conflicts of interest line, oxymetazoline, pseudoephedrine When a drug is needed to relieve sore ©Prescrire and tuaminoheptane) can cause seri- throat, paracetamol is a better option, ous and even life-threatening cardio- when taken at the appropriate dosage. Selected references from Prescrire’s literature vascular disorders, including hyper- search. 1- Prescrire Editorial Staff “Towards better patient tensive crisis, stroke and arrhythmias. care: drugs to avoid in 2015” Prescrire Int 2015; 24 This is unacceptable for drugs that are Putting patients first (158): 78-79. indicated for minor, rapidly self-­ 2- Prescrire Editorial Staff “Towards better patient care: drugs to avoid” Prescrire Int 2013; 22 (137): resolving ailments such as the com- Our analyses show that the 108-111. mon cold (Prescrire Int n° 136). harm-benefit balance of the drugs list- 3- Prescrire Rédaction “Des médicaments à écarter • Omalizumab, an anti-IgE monoclo- ed here is unfavourable in all their pour mieux soigner: pourquoi?” Rev Prescrire 2013; nal antibody approved in severe per- authorised indications. Yet some have 33 (360): 792-795. 4- Prescrire Editorial Staff “Determining the sistent asthma and chronic sponta- been marketed for many years and are harm-benefit balance of an intervention: for each neous urticaria, causes disproportionate commonly used. How can one justify patient” Prescrire Int 2014; 23 (154): 274-277. adverse effects: infections, hypersensi- exposing patients to drugs that have 5- Prescrire Rédaction “Objectifs des traitements: à partager avec les patients” Rev Prescrire 2012; 32 tivity reactions and cardiac disorders more adverse effects than other mem- (345): 544-546. (Prescrire Int n° 115, 161). Cortico­ bers of the same pharmacological class steroid therapy at the lowest effective or other similarly effective drugs? And

New Products Reviews

– Dasabuvir in hepatitis C – Ranibizumab or bevacizumab in age-related – Dalbavancin in bacterial infections of the skin macular degeneration – Ramucirumab in gastric cancer – Pivmecilinam in cystitis

Adverse Effects Outlook

– Ondansetron and pregnancy: possible congenital – New drugs and indications in 2015 heart defects – United States: student action reduces industry – Drug-induced hair loss influence in medical schools – Bradycardia with sofosbuvir

Prescrire International April 2016/Volume 25 N° 170 • Page 111