New products necessitating blood transfusions 3- “Paclitaxel” Prescr Intern 1994; 3 (14): 164-165. Literature 4- Rowisky EK et al. “Phase I and pharmacolog- occurred in 37% of patients (9). ic study of topotecan: a novel topoisomerase I The most frequent non haematologi- Our literature search was based on systematic inhibitor” J Clin Oncol 1992; 10: 647-656. cal side effects were nausea (68.1% of scrutiny of contents listings of the main inter- 5- Verweij L et al. “Phase I and pharmacokinet- ics study of topotecan, a new topoisomerase I patients, severe in 6.1% of patients), national journals and Current Contents at inhibitor” Ann Oncol 1993; 4: 673-678. vomiting (44.3%, severe in 4.5%), hair the Prescrire library, and on reference texts 6- Saltz L et al. “Phase I clinical and pharmacol- in clinical pharmacology (Martindale The ogy study of topotecan given daily for 5 consecu- loss (56.9%), fatigue (44.5%, severe in Extra Pharmacopoeia 31st ed., etc.). We also tive days to patients with advanced solid tumors, 6%), diarrhoea (26.1%, severe in 3.4%), consulted CD-ROM versions of Medline (1981- with attempt at dose intensification using recom- and stomatitis (20.2%, severe in 2%) (9). March 1998), Embase Drugs and binant granulocyte colony-stimulating factor” J Pharmacology (1991-January 1998), Cochrane Natl Cancer Inst 1993; 85 (18): 1499-1507. Paclitaxel solution contains a solvent, 7- Kudelka AP et al. “Phase II study of intravenous Cremophor EL°, that is incompatible with (1998, issue 1), Medidoc (1991-1994) and topotecan as a 5-day infusion for refractory epithe- Reactions (1983-June 1997), and the Minitel lial ovarian carcinoma” J Clin Oncol 1996; 14: the use of PVC infusion devices and war- version of the Pascal database up to 1552-1557. rants prior steroid administration because September 16 1997. SmithKline Beecham 8- Creemers GJ et al. “Topotecan, an active drug of its high sensitising potential (3). This is provided us with published and unpublished in the second-line treatment of epithelial ovarian documents, including the clinical expert cancer: results of a large European phase II study” not the case with topotecan. J Clin Oncol 1996; 14: 3056-3061. report. We also used the European public In all, 0.9% of patients died from 9- Ten Bokkel Huinink et al. “Topotecan versus assessment report (EPAR) available from the proven or probable toxicity on topote- paclitaxel for the treatment of recurrent epithelial European medicines agency. ovarian cancer” J Clin Oncol 1997; 15: 2183-2193. can, an incidence similar to that report- 10- European Agency for Medicinal Products ed with paclitaxel (9,11). And 5% of 1- NIH Consensus Development Panel on Ovarian (EMEA) “European public assessment report Cancer “Ovarian cancer - Screening, treatment (EPAR) Hycamtin” November 12, 1996: 32 pages. patients discontinued treatment with and follow-up” JAMA 1995; 273: 491-497. 11- Trimble EL et al. “Paclitaxel for platinum-refrac- topotecan because of severe adverse 2- Van Der Burg MEL et al. “The effect of debulk- tory ovarian cancer: results from the first 1 000 events (9). ing surgery after induction chemotherapy on the patients registered to National Cancer Institute prognosis in advanced epithelial ovarian cancer” Treatment Referral Center 9103” J Clin Oncol ©PI N Engl J Med 1995; 332: 629-634. 1993; 11: 2405-2410.

Translated from Rev Prescr Dec 1997; 17 (179): 807-809 benfluorex: what use? A SECOND LOOK POOR ASSESSMENT FILE

Abstract been done with morbidity or mortality contradictory. None of the three trials as end points. Available placebo-con- comparing benfluorex with a fibrate is ● The clinical file on the value of ben- trolled trials are small, methodologi- interpretable, because of fluorex for hypertriglyceridaemia is cally weak in most cases, and their methodological problems. highly inadequate. No clinical trials have results regarding triglyceride levels are

NOTHING There is no reason to prescribe benfluorex for hypertriglyceridaemia and diabetes. NEW

enfluorex (Biopharma, Servier most important question is the following: Tablets group), has been marketed in does benfluorex have proven • 150 mg per tablet B France since 1976, as tablets con- taining 150 mg (a). Chemically, benfluo- Biopharma a-In Europe, benfluorex is also marketed in Spain, rex is similar to other appetite suppres- Greece, Italy, Luxemburg, Portugal and ■ licensed indication in lipid disor- sants (b)(1). In France benfluorex has Switzerland (1). Benfluorex is not marketed in ders two approved indications. We recently English-speaking countries or Northern Europe. b- In France benfluorex has a peculiar classifica- “Adjunct to appropriate dietary mea- took another look at the file on the indi- tion. It is not officially classified among the appetite sures in hypertriglyceridaemia. The cation worded “adjunct to dietary mea- suppressants, meaning that its prescription is not restricted. However, the authorities decided on dietary measures must be continued. sures in asymptomatic overweight dia- October 25, 1995 (Journal Officiel, October 31, NB. There is no proven efficacy in pri- betic patients”, and concluded that its 1995: 15 937) to place benfluorex on a list of sub- mary or secondary prevention of com- stances that must not be used in freshly prepared slight effects on surrogate end points in mixtures, a list composed only of appetite sup- plications of atherosclerosis.” no way warranted prescription to diabetic pressants. Furthermore, the suffix -orex is attrib- uted to international nonproprietary names of lipid-lowering drug? patients (2). appetite suppressants by the World Health The second approved indication for Organisation (WHO) (1), and benfluorex is on the list of doping substances, together with amphetamines benfluorex is in the treatment of hyper- and other stimulants (Dictionnaire Vidal, French triglyceridaemia (see inset opposite). The data sheet compendium, 1997 edition, page 6).

P RESCRIRE I NTERNATIONAL A PRIL 1998/VOLUME 7 N° 34 • 49 Downloaded from english.prescrire.org on 28/09/2021 Copyright(c)Prescrire. For personal use only. New products / benfluorex

beneficial effects on the only clinical- treatment, five trials showed a statisti- f- The report does not state whether the investi- ly relevant end point, i.e. cardiovascular cally significant difference in triglyceride gators or the patients were blinded. g- The patients weighed at least 65 kg, in which morbidity and mortality, in patients with or levels between benfluorex and the place- case the recommended dose of clofibrate is without a history of cardiovascular dis- bo, while five showed no such difference. 2 g/day. ease? h- The authors themselves stated that the prob- Trials versus other lipid-lowering ability of a response to pharmacological treatment A consensus conference in the United drugs: uninterpretable results. Three was higher in the case of type IV dyslipidaemia States was devoted to hypertriglyceri- trials comparing benfluorex with other than in type IIB dyslipidaemia. daemia (3). According to the conclusions, lipid-lowering drugs have been published only marked hypertriglyceridaemia war- in detail (12-14). Two of these trials com- rants treatment, with the aim not only of pared benfluorex with clofibrate in reducing the hypertriglyceridaemia, but patients with hypertriglyceridaemia (at Literature more importantly of reducing LDL-cho- least 1.5 g/l) and/or hypercholestero- lesterol levels and increasing HDL-cho- Our literature search was based on system- laemia (cholesterol level above atic scrutiny of contents listings of the main lesterol levels. This is because the link 2.5 g/l) (e). The first trial was single-blind international journals and Current Contents between cholesterol fractions and car- and involved 40 patients (f)(12). While at the Prescrire library, and on reference diovascular morbidity is better estab- total triglyceride levels fell more strong- texts in clinical pharmacology (Martindale lished than for triglycerides. First-line ly on benfluorex (52.7% versus 38.5% The Extra Pharmacopoeia, etc.). We also management of hypertriglyceridaemia is consulted CD-ROM versions of Medline on average; p<0.001), the result was (1966-January 1998), Embase Drugs and based on lifestyle measures (giving up meaningless for at least two reasons: Pharmacology (1991-December 1997), smoking and alcohol, taking exercise, firstly, 15 of the 20 patients received only Reactions (1983, December 1997), and etc.) and dietary advice (increasing oily 1.5 g of clofibrate daily, an inadequate Cochrane (1998, issue 1), and the Minitel fish consumption). According to the version of the Pascal and EMC databases, up dose (g); secondly, patient recruitment to October 18, 1997. Servier sent us pub- American consensus conference, drugs was highly heterogeneous, the clofibrate lished and unpublished documents. Two are useful only when these measures group comprising more patients with type ISDB members (International Society of are inadequate (3). IIb hyperlipidaemia and fewer patients Drug Bulletins) - Boletin Terapeutico Lipid-lowering drugs known to induce Andaluz and Informazion sui Farmaci - sent with type IV hyperlipidaemia than the us their documentation. a marked reduction in hypertriglyceri- benfluorex group (h). daemia are the fibrates, nicotinic acid The second trial, involving 28 patients, 1- “Benfluorex°” Micromedex 31/03/1996; 87: 9 and fish oils (c)(3,4). Gemfibrozil, a pages. is also uninterpretable: blinding was not 2- “Benfluorex antidiabetic?” Prescr Intern 1997; 6 fibrate, is the only drug having shown a mentioned and this leads us to believe (30): 108-109. preventive action, based on clinical cri- that it was not done; the daily dose of 3- NIH consensus development panel on triglyc- teria, in a trial involving patients with eride, high-density lipoprotein, and coronary heart clofibrate was probably inadequate for a disease “Triglyceride, high-density lipoprotein, and hypertriglyceridaemia (d)(3,5). number of patients; and the distribution coronary heart disease” JAMA 1993; 269 (4): 505- We examined the clinical file on ben- of the various types of dyslipidemia was 510. fluorex for answers to the following ques- 4- “Lipid regulating agents”. In: “Martindale - The not clearly stated (13). Extra Pharmacopoeia” 31st ed., The Pharmaceu- tions: Does benfluorex have a proven The third trial involved 28 patients with tical Press, London 1996: 1299-1302. effect on cardiovascular morbidity? and 5- Hulley SB and Avins AL “Asymptomatic hyper- type 2 diabetes treated for 1 month with triglyceridaemia” BMJ 1992; 304: 394-396. Has it at least been compared with other benfluorex (450 mg/day), fenofibrate 6- Miller DS et al. “A study of the energy and bio- triglyceride-lowering drugs on the basis (300 mg/day), or bezafibrate (600 mg/ chemical status of obese and non-obese students of surrogate end points, i.e. triglyceride treated with 780SE” Postgrad Med J 1975; 51 day) (14). The paper mentions no sta- (suppl. 1): 117-120. and triglyceride-rich lipoprotein levels tistical analysis of intergroup differences 7- Institut de recherches internationales Servier (VLDL and chylomicrons)? “S780 - benfluorex - brochure pour investigateur” and, in any event, it seems impossible version n°1, May 30, 1996 (unpublished); 61 to show significant differences between pages. such small groups. 8- Krzentowski G et al. “Effet du benfluorex sur le Poor-quality, generally devenir métabolique d’une surcharge glucosée uninterpretable clinical trials There are no clinical trials of benfluo- orale chez le patient obèse avec tolérance au glu- rex using clinical end points in patients cose diminuée” Thérapie 1979; 34: 445-455. with hypertriglyceridaemia. 9- Ranquin R “Effects of benfluorex on patients The clinical file is of poor methodolog- with endogenous hypertriglyceridemia” Curr Med ical quality. In a recent article we underlined the Res Opin 1987; 10: 521-526. lack of clear and independent data on 10- Di Martino G et al. “Effects of benfluorex in Placebo-controlled trials: no con- the adverse effects of benfluorex (1). We obese patients with metabolic disorders” Br J Clin vincing evidence. Benfluorex has been Pract 1989; 43 (6): 201-208. were unable to find any new information 11- Bianchi R et al. “Effects of benfluorex on insulin compared with a placebo in about ten to contradict this judgement. resistance and lipid metabolism in obese type 2 diabetic patients” Diabetes Care 1993; 16 (4): 557- clinical trials, most of which have been ©PI published, but often vaguely descri- 559. 12- Balestreri R et al. “Effet thérapeutique com- bed (6-11). The trials involved between paré du benfluorex et du clofibrate dans les trou- 7 and 50 patients with various conditions bles métaboliques” Gaz Med de France 1982; 89 c- hMG-CoA reductase inhibitors (statins) have (14): 1636-1644. (glucose and lipid overload, hyperlipi- mild effects on triglyceride levels. 13- Di Perri T and Guerrine M “Etude compara- daemia, obesity with or without carbo- d- In the “Helsinki heart study”, although gemfi- tive du benfluorex et du clofibrate dans les hyper- hydrate disorders) treated in parallel brozil reduced triglyceride levels by 35% its pro- lipoprotéinémies de type IIa, IIb, IV” Acta tective effect on the cardiovascular risk was attrib- Therapeutica 1981; 7: 335-343. groups or cross-over studies, for 7 to 50 uted to its actions on cholesterol fractions (ref 5). 14- Sommariva D et al. “Differential effects of ben- days. The results are difficult to interpret, e- In these trials the patients were treated in par- fluorex and two fibrate derivatives on serum as most of the studies have major allel groups for 2 months with either benfluo- lipoprotein patterns in hypertriglyceridemic type 2 rex (450 mg/day in 3 doses) or clofibrate diabetic patients” Curr Ther Res 1996; 40 (5): 859- methodological biases. At the end of (1 500 mg/day in 3 intakes). 870.

50 • P RESCRIRE I NTERNATIONAL A PRIL 1998/VOLUME 7 N° 34 Downloaded from english.prescrire.org on 28/09/2021 Copyright(c)Prescrire. For personal use only.