arznei-telegramm 2001; 32: 88-9



Scandals with drugs such as LIPOBAY appear to follow a rather uniform script: There are public warnings (a-t 2001; 32: 36). The producer - after some delay - takes the drug off the market. The drug regulatory authorities remain inactive. This has been the case with the antabuse-like drug nitrefazole (ALTIMOL; a-t 1984; no 6: 45-6) used in alcohol dependence, with the antidepressant nomifensine (ALIVAL; a-t 1991: no 10: 93-7), with terodiline (MICTROL; a-t 1991; no 8: 72) for urinary retention and many others. More than three weeks passed after withdrawal of terodilin from the market before the regulatory authorities showed any reaction at all.

In the LIPOBAY case alike, the producer was first to act, but lost all credibility by incompetent risk management: Share-holders were the first ones to be informed of the market withdrawal which was published in the press before pharmacists and doctors were informed.

This was followed by the usual "who is the bad guy" procedure of the bureaucrats. The ministry of health declared that it was going to fine Bayer for not reporting a study on adverse events, but this was only window-dressing to divert from its own failure to act. Bayer had presented new risk data to the responsible licensing authority, the MCA, in the UK, in June. These data were presented to the responsible European Committee on Proprietary Medicinal Products (CPMP) on June 27, 2001. Neither the German nor any other national regulatory authority demanded market withdrawal of LIPOBAY.

The events are symptomatic for the worldwide decline in the quality of the drug licensing procedures (a-t 2001; 32: 57-8) and for the lack of an efficient system to monitor drug safety after marketing authorization. These structural deficits in drug approval and safety cannot be overcome by hectic activities such as calling for the immediate introduction of a questionable chip-card for patients. More fundamental changes are necessary and overdue:

  Slow-down of licensing: Drug approval procedures have around the world been accelerated following pressure of the industry for profitable fast market authorization. The regulatory authorities do no longer compete for highest quality of drug safety controls, rather for highest efficiency in granting licenses (a-t 1998; no 4: 37-8). Inefficient and unsafe drugs pass the hastened licensing procedures more easily. The risks in drug safety are transferred from the producer to the patients who become the guinea pigs for fast marketed, but insufficiently tested drugs. A responsible regulatory authority needs at least one or two years of thorough testing to decide on a new therapeutic principle.

  Abandon surrogate criteria: Surrogate end points such as reduced cholesterol levels or lowered blood pressure do not suffice to evaluate drug efficacy (a-t 1999; no 11: 113-5). Clinical studies are needed with "hard" clinical end points such as heart failure or cerebrovascular events to decide on licensing and to prevent errors. For instance, the US regulatory authority FDA denied license of alprostadil (PROSTAVASIN) approved and marketed for treatment of peripheral vascular disease in Germany, since the data indicated no effect on the number of amputations caused by the disease, but only a questionable benefit in intermittent claudication as surrogate end point (a-t 1995; no 12: 113). The risk to be deceived by surrogate criteria applies to all indications. In oncology, surrogate criteria such as "time-to-progress" are without meaning for the evaluation of the efficacy of a cytostatic drug, if it does not prolong survival or improve quality of life.

  Drugs used for prolonged treatment require long-term studies: Drugs for long-term treatments often develop a delayed toxic potential. Therefore, safety can only be evaluated with data from long-term studies, which must be done before approval and must be presented for licensing. If there are indications that potential risk factors of the indicated diseases are enhanced - e.g. increased blood pressure with the anti-obesity drug sibutramine (REDUCTIL; a-t 2001; 32: 27) or increased weight by the anti-diabetic glitazones (a-t 2001; 32: 64) - these drugs should not be licensed without clinical data on the respective long-term risks.

  Mandatory publication of all licensing data: Decisions of regulatory authorities must be verifiable. Therefore, all relevant clinical studies and all safety data have to be available to the scientific community at the time of licensing in order to prevent secret deals between the applicants and the officials. In addition, only this provides a chance to recognize selective or manipulated publications, which are in contrast to data forwarded to the licensing authorities (a-t 2001; 32: 87-88).

  Monitoring for drug safety by a pharmacovigilance-system: Rare, but serious adverse drug reactions are most often recognized only after marketing of a drug, when many patients have already been exposed. In Germany, we expect about 210.000 serious drug-induced diseases per year requiring hospitalization, out of which 70.000 are highly critical requiring intensive care treatment and 16.000 are fatal. (1) The regulatory authority BfArM receives only 5.000 to 7.000 reports of serious adverse events per year from the medical profession or the industry. The office responsible for pharmaceutical risk prevention only receives notice of about 3% of all adverse reactions. Therefore, risks are realized too late and measures to reduce those risks are introduced too late. As a the consequence avoidable severe drug-induced diseases and fatalities occur.
A pharmacovigilance system consisting of 10 centers continuously monitoring about 5% of the hospital admissions in Germany would be capable of recognizing potential drug risks earlier and faster. Such a system costs about 5 million Euro annually. These expenditures would be more than balanced by the reduction in the number of drug-induced diseases prevented by such a system. International data show that 30% to 50% of all drug-induced events are avoidable. (2, 3) 210.000 severe drug-induced diseases demand about 2 billion Euro annually for treatment costs according to data from the US. The improved and accelerated recognition of adverse events would, therefore, save about 1 billion Euro annually.

  Disclosing links to the industry: Drug information is the domain of the producer, either through sponsored or advertisement dependent medical journals (a-t 1998; no 8: 69-70), through drug representatives, by paid opinion leaders or by speakers for medical conferences with honoraria from the industry (a-t 1999; no 3: 33-4). Employing an apparent independent organizer as "impresario" often purposely conceals the dependence of such medical meetings on industry money and influence.
Therefore, it is necessary that sponsors have to be revealed in all publications and during all presentations at conferences. Direct or indirect honoraria as well as other financial support (paid expert opinions, shareholding etc.) have to be disclosed completely in order to gain transparency on the influence of producers on the drug information supplied.
Recently, the editors of 12 international "Peer-Review" journals set a signal: Because of increased pressure of the industry on publications it now becomes essential to declare which part authors and companies have had in the design, performance and evaluation of the data. (4)
In addition, legal basis has to be established to prosecute false or misleading information. It looks as if scientific honesty can only prevail, when malinformation based on marketing principles or manipulation of data is prosecuted by penal orders.

  Making warnings effective: Where is the effect of an official warning issued by the Medicine's Committee of the German Medical Profession in June, but not published before the end of August because of ongoing legal hearings? (5) Agreements with the industry to synchronize statements result in a blockade of information under the given condition. In addition, the medical profession has to regard warnings as binding for their own practice of drug prescribing.


Consequences have to be drawn from the LIPOBAY affair. The next candidates for market withdrawal can be pointed out right now, having been registered under decreased standards of drug approval and safety surveillance: the anti-diabetic glitazones termed "insulin-sensitizers".

Licensing has to be based on thorough investigations without time pressure.
Surrogate criteria must be excluded from assessment of drug efficacy.
Long-term clinical studies are mandatory for licensing long-term use of a drug.
Continuous monitoring of drug safety has to be established by a post-marketing pharmacovigilance-system.
Drug information must be free of hidden influences of the drug industry at all levels of professional training of physicians.
Links of experts or opinion leaders to the pharmaceutical industry and financial relationships have to be disclosed during professional meetings and - finally in Germany as well - in medical journals.
Manipulation of data in favor of the producer has to be prosecuted and punished.


SCHÖNHÖFER PS et al.: DGPT-Forum 2001; no. 28: 15-9


Bates D.W. et al.: JAMA 1997; 277: 307-11


CLASSEN D.C. et al.: JAMA 1997: 277: 301-6


DAVIDOFF F. et al: Lancet 2001; 358: 854-6


MÜLLER-OERLINGHAUSEN B.: Südd. Ztg. Aug. 25, 2001

© arznei-telegramm 09/01