Off label use—label off use?

J. Boos

University Children’s Hospital Muenster, Department of Paediatric Haematology and Oncology, Münster, Germany (E-mail: boosj@uni-muenster.de)

During the last century broad medical progress arose from the development of treatment strategies from natural remedies and the industrial production of drugs. Where there used to be a direct relationship between the natural healer or physician recommending an individualised treatment, the pharmacist preparing a custom-made prescription and the single patient, there is now a drugs market with the pharmaceutical industry on the supplier side, the patient as consumer and physicians who ideally prescribe drugs based on evidence and according to pertinent guidelines. The reproducibility and quality of diagnostic procedures as well as the drug supply have increased and the general benefit for society and individual patients is obvious. Cancer patients have benefited significantly from this process and the overall cure rate of >70% in paediatric malignancies would not have been achieved without cytostatic drugs developed and provided by the pharmaceutical industry.

This process, however, was paralleled by a number of dramatic setbacks. Heroin was sold for years, diethylstilbestrol induced cancer in next generation girls, chloramphenicol caused newborns to die from Grey’s syndrome, coagulation factor preparations infected patients with HIV, to list a few examples. And no one will forget the hundreds of babies born with malformations induced by thalidomide use during pregnancy. It became evident that uncontrolled availability of chemical substances and plant poisons marketed as medicines, on the basis of developmental hypotheses, providers’ beliefs or even limited observation of effects, produced a risk for the population.

As a result, governments developed specific drug laws and most societies try to control the drugs industry in this way. Based on such legal guidelines, it is mandatory that critical drugs are prescribed by physicians, and that pharmaceutical companies provide preclinical efficacy and safety data in animals as well as results of clinical trials to prove the safety profile and the efficacy for the intended indication of a new drug going onto the market. Thus, the drug label was born. And nobody will question the aims of labelling: pharmaceutical quality, medical evidence, sufficient information on side-effects, known risk–benefit ratio, readable information for the user, consumer protection, etc.

If, however, we have accepted society’s general need to control the drug market and have learned the historical lesson that especially the developing organs of children are at high risk from systematic drug toxicities, why then do children with leukaemia have to be treated with unlicensed or off label drugs as Conroy et al. report in this issue [1]?

When a drug is approved, indications and dosages are part of the package insert. It is only for these particular situations that the risk–benefit ratio has been reviewed and accepted by the regulatory authorities. In any other circumstance, the drug is used outside the limits of its label, with the result that neither the company nor the authorities take any legal or ethical responsibility for unexpected events. Different dosages, use outside the labelled indications, modifications of the pharmaceutical form (breaking up tablets to prepare capsules with smaller amounts of the drug or making solutions for children are classed as ‘special formulations’) or of the route of application (i.v. solutions for oral or intrathecal use), many such changes will mean that use of the drug is without the formal and legal protection of the label—off label use.

The paper by Conroy et al. comprehensively summarises and discusses the field of off label use in children [1]. A huge number of papers have focussed on this problem in recent years [27]. In summary, the younger the patient and the more critical and rare their illness, the more likely they will need treatment off label. Paediatric intensive care units and oncology wards have much higher proportions of off label use than general paediatric practice, where ~80% of the drugs are reported to be used in accordance with their label [4, 8].

The phenomenon of off label use, however, is neither restricted to highly specific clinical situations nor to single countries. Off label use is an international problem with comparable proportions reported in European countries and North America. In children, all classes of drugs are involved. Even ~60% of important antidotes used in the treatment of poisoning have to be prescribed off label in children [9]. In Germany, we have recently reviewed the availability of essential drugs according to the 11th WHO’s Essential Drug List [10]. In 17 selected therapeutic groups (not groups of physiological substances such as vitamins, minerals, blood products) ~30% of essential drugs were contraindicated for children in Germany (unpublished data). If >50% of treatment with any drug happens off label, the requirements of the regulation are not met. The label is of no use in actually regulating the market of the drug.

Apart from paediatrics, off label use evidently plays an important role in adult oncology. New classes of drugs enter the market for very specific indications with narrow therapeutic goals. Gemcitabine was initially labelled for progressive pancreatic cancer, paclitaxel and topotecan for second-line ovarian cancer. Once on the market the number of labelled indications increases for some drugs: today paclitaxel is labelled for ovarian cancer, breast cancer, Karposi’s sarcoma and non-small-cell lung cancer; gemcitabine has additionally been labelled for non-small-cell lung cancer; and topotecan for small-cell lung cancer. All of these drugs, however, are also in clinical use in many other indications.

In children, the off label problem is often exacerbated by the fact that the drugs not only have to be used off label, but against specific warnings and contraindications. Most carboplatin package inserts list children among the contraindications. Nevertheless, >200 studies of this drug in children have been published, and >40 pharmacokinetic investigations including population based pharmacokinetic trials have been conducted. Roughly half of all children with cancer in Germany currently receive carboplatin at some time during their treatment. This clearly shows that the labelling neither represents the clinical evidence nor follows it.

If, as Conroy et al. report, all one’s patients receive drugs off label and one is routinely forced to prescribe broadly accepted and evidence-based drug treatment strategies despite dramatic warnings in the package leaflets, such unsubstantiated warnings undermine the value of labelling in general. Who will decide whether or not warnings are substantiated? Who will explain the discrepancy between reality and leaflets to patients and parents over and over again? The principal consequence arising from the need for routine off label use is merely that the label in general is off use in the physician’s decision making. This, however, may result in significant problems with drugs where warnings need to be heeded!

Labelling, scientific knowledge, medical evidence and clinical experience are different sides of the same coin: the true value of a new drug for the patient population. Regulatory authorities can only regulate the licensing process and the label, they cannot regulate the practice of medicine [11]. In Germany and probably in other countries as well, this part of ‘regulation’ has led to the question of whether or not the health system should pay for drugs used off label. If not, the prescribing physician not only takes on responsibility for the drug, he or the patient might eventually have to take over the cost. State programmes and insurances deny reimbursement for drugs prescribed for off label indications. The arising conflict for physicians and patients is being discussed in the USA as well as Europe [12, 13].

Labelling, initially thought to guarantee the consumer a proven safety–efficacy relationship, now tends to be used to regulate economic resources in health systems.

It is up to the physician to offer the best available treatment to the patient, it is up to society (as a public health system or via health insurance) to provide the financial resources for each patient’s treatment, it is up to the regulatory authorities to protect patients from dubious drugs and inadequate marketing. The whole responsibility for supplying drugs, however, in the current system is transferred to the pharmaceutical industry. Only drug companies are in a situation to apply for a label or label changes. But why should they? Why should they label carboplatin if it is widely used anyway? Why should they label sulfamethoxazol for prophylaxis of Pneumocystis carinii infections in immunocompromised children if its use is obvious? Why develop low strength mercaptopurine tablets when it has been off patent for years?

The pharmaceutical companies’ greatest interest lies in coming to market as soon as possible in order to have maximum benefit from remaining patent protection and to refinance the high cost of drug development, and at last to earn money. Today, most pharmaceutical companies are international stock companies whose primary aim is to increase profit and shareholder value. They expand, employ people, pay taxes, and are bound by all laws regulating business. Investment and earnings have to be balanced for companies. In a situation where licenses are issued for a specific indication and the market is open for additional indications, the primary use of a label for the company is to make the drug initially available on the market. All decisions about label changes and additional indications will be based on calculations regarding investigation and earnings in the new fields. Paediatric cancers, as well as rare malignancies in adults, thus had little chance of being part of the post-marketing development, until the US Food and Drug Administration (FDA) changed its policy and demanded paediatric development plans.

Today, the regulatory authorities of different countries cooperate closely and try to standardise the demands for labelling. The International Conference on Harmonisation agreed on Good Clinical Practice guidelines (ICH-GCP), which define standard procedures and requirements for clinical trials in drug development [14]. These guidelines address quality, safety and efficacy as primary goals and intend to guarantee that the labelling of a drug will be based on valid, true and unbiased sets of data, and that the maximum safety and medical standards will be guaranteed for participating patients. These requirements play an important role as a counterbalance to the business interests of industry. In addition, however, they increase the cost of clinical trials in drug development to enormous amounts; this is not only true for industry-sponsored trials but also for investigator-initiated trials. The European Commission estimates the cost of a paediatric drug trial conforming to GCP guidelines to be $5–10 million [15], an enormous amount of money for a population representing <1% of all cancers.

Currently, it is the pharmaceutical company who decides which indication will be the focus of development and label application. To be on the safe side, not risking any delay in licensing, the indication will be chosen precisely and well defined, in a tumour entity that promises to respond to chemotherapy and represents a big market: relapsed breast cancer for example.

The authorities cannot delay the licensing of a drug of proven benefit in the defined target patient population. Once marketed, however, the new drug will be available to everybody and will be used off label (and without limits). Especially when the mechanism of action seems to be new and broad ranging—as is the case with taxanes, topoisomerase I inhibitors and some other classes of drugs—nearly every cancer patient will have a realistic chance to benefit from the drug. Parents will reliably ask for new drugs they learned about on TV or in internet chat rooms. They try to grasp the new chance for their child, and do not want to watch and wait and possibly realise years later that all along the drug had been on the market and was eventually proven to be effective in their child’s condition—but had not been labelled accordingly before their own child had died. Adults in a palliative situation have similar feelings: they want access to new drugs immediately and cannot wait for the long process of scientific evidence formation and labelling to run its course. This is why parents and patients actively and intensively ask for clinical trials with new drugs as well as experimental strategies.

The major political and regulatory task in this clear conflict of interest takes us back to the root of the problem: how to offer new drugs to the patients as early as possible while controlling the risk–benefit ratio and protecting those patients from simple profiteering whose only option is hope.

Testing indication by indication and age group by age group in various combination protocols as a precondition for comprehensive labelling cannot work in rare diseases. In a country like Germany no more than 20 patients per year qualify for experimental treatment strategies in malignancies such as Ewing’s tumour or neuroblastoma. In this situation, the amount of safety and efficacy data required to submit an application for drugs such as antibiotics, antihypertensives or even cytostatics cannot be obtained. We have to face the fact that the expectation to supply standard data sets with hundreds of patients is an illusion in severe paediatric illnesses and rare malignancies. Lower levels of significance or lower power might be worth thinking about, but ultimately nobody wants to systematically reduce the standards for specific patient groups.

As nearly all available cytostatic drugs show activity in a spectrum of malignancies, the definition of an ‘indication’ is also worth discussing. ‘Paediatric solid tumour’ or ‘embryonic tumour’ or ‘neuroblastoma’ or ‘MYCN-positive relapsed neuroblastoma’, which is an indication? A new drug will be used in all of these definitions, but trials become more difficult the more specific the malignancy to be included. The FDA Center for Drug Evaluation and Research states in a draft guideline for industry about paediatric oncology studies "...phase III pediatric oncology studies are generally the postapproval standard of care for children with various malignancies..." [16]. Herewith, the FDA accepts the phase III standard of the cooperative trials in paediatric oncology. The only question is, why don’t the results translate into label changes?

Current trials work on the basis of European or even worldwide cooperative study protocols and treatment development. The paediatric oncology societies have thus established a network for developing evidence-based treatment strategies, scientific development, evaluation of new drugs in a well organised context and, last but not least, quality control and patient protection. These trials may serve as an example for other rare diseases and for structures which are or should be organised for rare adult malignancies as well. The current problem for many of the coordinating investigators of these trials is, however, that they find themselves in a conflict between patient care, the requirements of a GCP-like drug development trial and lack of sufficient financial resources.

To better adapt the system of investigator-initiated phase III trials, the specified physicians and scientific societies should be heard when companies seek scientific advice. They might serve as catalysts or representatives of patients’ interests somewhere between regulatory claims and economic interests. At the same time, the oncological trial groups should also have the opportunity to recruit free-of-charge scientific advice. This could lead to small, but significant changes in protocol designs which could easily adapt them better to regulatory requirements. Then, the gap between published scientific knowledge and facts accepted by regulatory requirements that enter labelling might shrink.

Only round tables involving investigators, companies and regulators will be in the position to optimise the conflicts between small patient numbers; statistical efforts, degree of significance level and statistical power; and financial resources and financial needs for multicentre activation and monitoring. Provided, the participants agree on a compromise, in the end, this compromise must not be in patient safety or data quality. Eventually, the approval procedure should use the available resources and lead to a labelling that addresses knowledge and open questions, and allows the use of a drug in the multicentre treatment protocols of scientific and clinical societies.

From my point of view this is a crucial point: whenever something very complex is going to be used the party responsible for the safety of its use has two possible options either provide long and specified handling instructions or restrict the use to intensively trained specialists.

In all developed countries paediatric oncology treats children predominantly in well organised trials. Labelling that restricts the use of a drug to specialised physicians and to the context of GCP-adapted studies with appropriate quality assurance, combined with an effort by companies to annually report on drug use in paediatric tumour types, might open a way for cooperation between those companies as well as authorities and clinical trial groups. The well organised system of national and international (paediatric) oncological trials is still developing along with the increasing complexity of GCP guidelines.

To overcome the problem of off label use, constructive cooperation between companies, authorities and clinicians in such a post-marketing surveillance system, with the companies taking on principal responsibility for the drug, offers a number of opportunities. The system of clinical phase III trials and the overhead cost of treatment protocols would logically become part of the regular health system and might be supported by the pharmaceutical companies being responsible for reporting to the authorities. There is really no reason that quality control of standard care, clinical trials and drug development for rare diseases has to be financed by charity organisations as it is today.

The primary character of treatment optimisation, standardisation and quality control of these trials would no longer be doubtful only because of systematic off label use of drugs. Companies and, more importantly, authorities could gain a much better insight into the databases regarding response parameters and adverse event reporting, and patients who read this labelling would ask for physicians cooperating in such clinical trials.

Eventually, the proposed way of legalising the phase III use of a drug (after obtaining the required information from phase I and a few phase II trials) in paediatric oncology as well as in rare tumour types in adults will increase the value of cooperation in clinical trials. This will enhance the general gain in scientific and clinical quality of treatment in patients who are outside the focus of industrial drug development and the mainstream of clinical and scientific developments.

Today the database based knowledge about side-effects of drugs such as methotrexate or carboplatin in children is in the hands of the leukaemia and osteosarcoma trial groups. They are the first to recognise increments of relevant side-effects and they have the experience to adapt treatment strategies to specific clinical situations such as renal or liver insufficiency, extreme weight changes or others.

During the last few years the FDA has continuously developed regulations in order to streamline drug development and to include paediatric drugs and orphan diseases in this process. Interestingly, one of the changes with the FDA Modernization Act of 1997 [17] was that pharmaceutical companies were allowed to disseminate articles from peer-reviewed journals about off label use, which had previously been forbidden. The idea behind this approach was that companies should provide information about uses not specifically approved by the FDA to physicians having the legal right to prescribe drugs off label. "These changes will allow physicians to make more informed prescribing decisions and may motivate pharmaceutical companies to do the clinical studies necessary to get these indications added to drug labelling" [18]. This regulation will probably be useful in several clinical situations and off label indications. In oncology and especially in paediatric oncology, however, the treatment strategies are sufficiently complex and the burden of side-effects so high that generally patients ought not to be treated in institutions without access to the important peer-reviewed journals. In a qualified health system the evidence base of cancer treatment does not depend on the distribution of review articles by the pharmaceutical industry (although they still have the responsibility to have all, even unpublished, information available).

The distribution of reviews, therefore, cannot substitute for labelling, and the idea that supporting information by reviews, which usually summarize trial results, might stimulate label-directed clinical trials is unrealistic in oncology. If reviewed information is convincing enough to argue in favour of off label use this information ought to be included in the label. And only then will the information also be included in patient information leaflets.

A widespread off label use of drugs in 50–100% of a patient population is a declaration of bankruptcy. In this situation the label is off use in that it leads to general ignoring of the package inserts. Patients do not find relevant information in their package, physicians have to take on more legal responsibility, authorities cannot control the market, companies do not know what happens with their drugs, and all have to be aware of a higher adverse drug reaction rate in off label settings [19].

Legal changes and regulatory activities in the USA and Europe are definitely developing the system in the right direction (for more details see [20]). The Orphan Drug Act [21] and related European regulations, as well as currently active committees and increasing cooperation with physicians [22] and patients, represent successful initiatives. The US Best Pharmaceuticals for Children Act [23] and the consultation document Better Medicines for Children by the European Commission [15] open perspectives and realistic options for the subproblem of paediatric developments in the future. These strategies, however, will never release the resources to re-evaluate the current scientific and empirical knowledge which is the basis of the broad off label use.

In order to get rid of the off label use within a realistic time frame, some issues will, therefore, have to be taken into consideration much more than is currently the case.

First, we have to review and eventually legalise the broad off label therapeutic experience (especially with regard to old off-patent drugs) for children, wherever possible. The pharmaceutical companies should be required to summarise published data, medical guidelines and handbook experience with respect to their drug product and submit this information to the authorities. It is much better, for patients, physicians, society and even the pharmaceutical companies to have widespread use of drugs and standard treatments within the label rather than off label if the lack of labelling does not represent a lack of knowledge but only a lack of formal development and reporting, and if there is long-term experience regarding safety issues.

Secondly, especially for cytostatic drugs with usually more general mechanisms of action, new drugs ought to be labelled for less specific indications and dose regimens. The circumstances under which new drugs will be used are extremely variable. Dosages depend on cumulative pretreatment, comedication or the individual patient’s physiological status. Generally, they will be used in multi-morbid patients, not in standardised treatments following the precise inclusion criteria of developmental trials. Why not label indications in a less specific way omitting unsubstantiated warnings while adopting the restriction to "use only in quality controlled phase III trials" or comparable definitions to ensure close observation, adverse event monitoring and quality control? In this way, the labelled use of the drug guarantees a well organised quality and safety control, which is much stricter than most of the public reporting systems and off label use will apply only to those therapists not cooperating in trials. They will then have to take on the responsibility for its use with open eyes.

Thirdly, (fully aware that this may be difficult to achieve in our legal systems) I propose that the medical societies should also have available some official procedure to initiate label changes. They should be able to summarise the evidence based on investigator-initiated trials and apply for acceptance of the results by the authorities, resulting in inclusion on the label as reported experience.

The general off label use of drugs is the death of the idea of regulation. There is no doubt that against the background of tremendous economic interests and with hundreds of ‘designed’ and ‘targeted’ drugs about to enter the market, we urgently need more, not less regulation. This, however, can only be achieved if all participants in drug development, licensing and use cooperate flexibly in a constructive form of partnership that can be catalysed, more than it is today, by physician and patient initiatives. Under such circumstances, off label treatment could, for the first time, become an exception and receive the necessary attention. The available labels will re-enter appropriate use for companies, physicians, authorities and, most important of all, patients.

J. Boos

University Children’s Hospital Muenster, Department of Paediatric Haematology and Oncology, Münster, Germany (E-mail: boosj@uni-muenster.de)

References

1. Conroy S, Newman C, Gudka S. Unlicensed and off label drug use in acute lymphoblastic leukaemia and other malignancies in children. Ann Oncol 2003; 14: 42–47.[Abstract/Free Full Text]

2. Turner S, Gill A, Nunn T et al. Use of "off label" and unlicensed drugs in paediatric intensive care unit. Lancet 1996; 347: 549–550.[ISI][Medline]

3. Turner S, Longworth A, Nunn AJ, Choonara I. Unlicensed and off label drug use in paediatric wards: prospective study. BMJ 1998; 316: 343–345.[Abstract/Free Full Text]

4. McIntyre J, Conroy S, Avery A et al. Unlicensed and off label prescribing of drugs in general practice. Arch Dis Child 2000; 83: 498–501.[Abstract/Free Full Text]

5. Conroy S, Choonara I, Impicciatore P et al. Survey of unlicensed and off label drug use in paediatric wards in European countries. BMJ 2000; 320: 79–82.[Abstract/Free Full Text]

6. Conroy S, McIntyre J, Choonara I. Unlicensed and off label drug use in neonates. Arch Dis Child Fetal Neonatal Ed 1999; 80: F142–F145.[Abstract/Free Full Text]

7. Turner S, Nunn AJ, Choonara I. Unlicensed drug use in children in the UK. Paediat Perinatal Drug Ther 1997; 1: 52–55.

8. Bücheler R, Schwab M, Mörike K et al. Off label prescribing to children in primary care in Germany: retrospective cohort study. BMJ 2002; 324: 1311–1312.[Free Full Text]

9. Lifshitz M, Gavrilov V, Gorodischer R. Off-label and unlicensed use of antidotes in paediatric patients. Eur J Clin Pharmacol 2001; 56: 839–841.[CrossRef][ISI][Medline]

10. Essential Drugs and Medicines Policy (EDM). WHO [on-line] 2002; http://www.who.int/medicines/ (November 2002, date last accessed).

11. Cohen PJ. ‘Off-label’ use of prescription drugs: legal, clinical and policy considerations. Eur J Anaesthesiol 1997; 14: 231–235.[CrossRef][ISI][Medline]

12. Henry V. Off-label prescribing. Legal implications. J Leg Med 1999; 20: 365–383.[ISI][Medline]

13. Burgardt C. Off-Label-Use of drugs in the GKV. Forum 2002; 17: 24–26.

14. International Conference on Harmonization. [on-line] 2002; http://www. ich.org (November 2002, date last accessed).

15. Better Medicines for Children. European Commission [on-line] 2002; http://pharmacos.eudra.org/F2/pharmacos/medchild.htm (November 2002, date last accessed).

16. Guidance for industry. Pediatric oncology studies, in response to a written request. Food and Drug Administration—Center for Drug Evaluation and Research [on-line] 2001; http://www.fda.gov/cder/guidance/3756dft.htm (November 2002, date last accessed).

17. Food and Drug Administration Modernization Act of 1997. Food and Drug Administration—Center for Drug Evaluation and Research [on-line] 1997; http://www.fda.gov/cder/guidance/105-115.htm (November 2002, date last accessed).

18. Reh M. Changes at FDA may speed drug approval process and increase off-label use. J Natl Cancer Inst 1998; 90: 805–807.[Free Full Text]

19. Choonara I, Conroy S. Unlicensed and off-label drug use in children: implications for safety. Drug Saf 2002; 25: 1–5.[ISI][Medline]

20. Steinbrook R. Testing medications in children. N Engl J Med 2002; 347: 1462–1470.[Free Full Text]

21. The Orphan Drug Act. U.S. Food and Drug Administration [on-line] 2001; http://www.fda.gov/orphan/oda.htm (November 2002, date last accessed).

22. Press release: 1st EMEA workshop with health professionals and academia on orphan medicinal products. The European Agency for the Evaluation of Medicinal Products [on-line] 2002; http://www.emea.eu. int/pdfs/human/comp/003402.pdf (November 2002, date last accessed).

23. Best Pharmaceuticals for Children Act. U.S. Food and Drug Administration [on-line] 2002; http://www.fda.gov/opacom/laws/pharmkids/contents.html (November 2002, date last accessed).





This Article
Full Text (PDF)
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (2)
Disclaimer
Request Permissions
Google Scholar
Articles by Boos, J.
PubMed
PubMed Citation
Articles by Boos, J.