Data from regulatory studies: what do theytell? What don’t they tell?

Regulatory approval of new drugs by authoritiessuch as the United States Food and Drug Admin-istration and European Medicines EvaluationAgency requires adherence to strict clinical studyguidelines, to satisfactorily demonstrate efficacyand safety. Regulatory studies are conductedduring phase II (therapeutic exploratory) andphase III (therapeutic confirmatory) of a drug’sdevelopment. For antiepileptic drugs (AEDs), anew agent is always initially studied as add-ontherapy to one or more other AEDs inrefractory, partial onset seizure patients. Althoughregulatory studies provide useful informationabout an AED’s efficacy, tolerability and ease ofuse, they are limited as they are conducted inprotocol-restricted patient populations using fixedstudy designs and dosing schedules for shorttreatment periods. Therefore their relevance toeveryday clinical practice is poorly defined (1). Thispaper will discuss the limitations of regulatorystudies and the importance of phase IV studies,long-term follow-up of regulatory studies and

audit data in the overall assessment of a drug’sclinical performance.

What do regulatory studies tell us?

Phase III regulatory studies are mandatory for theapproval of an indication and are usually random-ized, double-blind, placebo-controlled and multi-center and conducted in patient populations that aredefined in terms of age, state of health, seizure typeand severity. Such studies are specifically designed toprovide clear-cut evidence of a drug’s efficacy,tolerability and acute side effects and ease of use.As they have rigid protocols with defined dosingtitration schedules, patient compliance is usuallyhigh. Add-on phase III studies of AEDs – forexample, the phase III levetiracetam studies (2–5) –are often predictive of tolerability and safety.Monotherapy studies, conducted only after AEDsare shown to be safe and effective in refractorypatients, provide further information on a drug’sefficacy when used alone (6, 7).

Acta Neurol Scand 2005: 112 (Suppl. 181): 21–25All rights reserved

Copyright � Blackwell Munksgaard 2005

ACTA NEUROLOGICASCANDINAVICA

Ben-Menachem E. Data from regulatory studies: what do they tell?What don’t they tell? Acta Neurol Scand 2005: 112 (Suppl. 181): 21–25.� Blackwell Munksgaard 2005.

Phase III studies of antiepileptic drugs (AEDs) are specifically designedto satisfy strict regulatory criteria. As they are conducted in protocol-restricted patient populations over short treatment periods and employfixed study designs and dosing schedules, they are not fullyrepresentative of �real-life� clinical practice. Therefore, in order toprovide an overall assessment of clinical performance, regulatorystudies must be backed up by post-marketing clinical experience. PhaseIV studies provide information on a drug’s performance in a settingmore closely representing real clinical practice, with broader patientpopulations and a more flexible approach to individual treatment.Prospective long-term studies allow the determination of efficacy andsafety (and cost-effectiveness) over extended treatment periods; thesestudies and audit data provide a means of assessing idiosyncratic sideeffects, unusual interactions and the effects of an AED in rare patientgroups. By complementing regulatory evidence with real-life clinicalexperience, a comprehensive assessment of the risks and benefits of anAED can be made.

E. Ben-MenachemDepartment of Clinical Neuroscience, SahlgrenskaAcademy, Gçteborg University, Gçteborg, Sweden

Key words: antiepileptic drugs; audit data; lamotrigine;levetiracetam; long-term follow-up; phase III studies;phase IV studies; regulatory studies

Elinor Ben-Menachem, Department of Clinical Neuro-science, Sahlgrenska Academy, Gçteborg UniversityHospital, 413 45 Gçteborg, SwedenTel.: +46 31 342 1000Fax: +46 31 211552e-mail: elinor.ben-menachem@neuro.gu.se

21

What don't they tell?

The strict inclusion criteria used in phase IIIregulatory studies result in patient populationsthat are not representative of those encountered inreal clinical practice. For example, add-on phaseIII studies conducted in refractory patients mostoften only include patients aged between 16 and65 years, who have failed to respond to more thantwo AEDs, are currently receiving treatment with1–3 AEDs and have a seizure frequency of morethan three to four per month for partial seizuresand more than one per month for generalizedseizures; they usually exclude patients who havenot had status epilepticus within the previous6–12 months, those with serious psychiatric illnes-ses and those with serious medical conditions.Regulatory studies are also limited in terms of theirduration and therefore cannot provide informationon long-term efficacy and safety.

Efficacy in the �real world�

Phase IV (therapeutic use) studies are important toassess the performance of a drug in everyday clinicalpractice. They are conducted in patient populationsthat aremore inclusive than those used in regulatorystudies, in terms of patient age, seizure type andseverity. They allow flexible dose adjustment for

individual needs, providing valuable information ontheoptimal use of a drug in away that is closer to realclinical practice. For example, Safety ofKeppra� asAdjunctive Therapy in Epilepsy (SKATETM) andKeppra�Epilepsy Evaluation of the Patient timE toResponse (KEEPERTM) were large phase IV levet-iracetam studies, designed to evaluate the safety ofadjunctivetherapywithlevetiracetamincommunity-basedpractice and toassess various aspects of its use,includingefficacy, inbroaderpatientpopulations (8–10). These 16-week, open-label studies includedpatients treated with a wide variety of concomitantAEDs who were experiencing persistent seizures, aswell as those with a number of comorbid medicalconditions; therewasnoupperagelimit for inclusion.Levetiracetam was initiated at 1000 mg/day andtitrated up to 3000 mg/day to access optimal clinicalbenefit.The efficacyof levetiracetam, in termsof rateof response and the proportion of patients whobecome seizure-free, was shown to be greater in theSKATETM and KEEPERTM studies (8–10) than inthe phase III regulatory studies (2–6) (Fig. 1). Addi-tionally, SKATETM and KEEPERTM were able todemonstrate that levetiracetam was effective acrossall partial seizure types, and that adverse events weremild to moderate in severity, rarely leading todiscontinuation (8–10).The disadvantages of phase IV studies are that,

as they are open-label and uncontrolled, patient

A

B

Figure 1. Comparison of the efficacy of levetiracetam in phase III regulatory studies (2–6) and phase IV studies [SKATETM (Safetyof Keppra� as Adjunctive Therapy in Epilepsy) and KEEPERTM (Keppra� Epilepsy Evaluation of the Patient timE to Response)](8–10), showing results for (A) 50% responder rates and (B) seizure freedom rates.

Ben-Menachem

22

compliance is often lower than in regulatorystudies and there is a higher risk of discontinu-ation. Also, they are inherently more biased as theinvestigator and patient know the medication theyare taking. Changes in concomitant medicationduring the course of the trial might also confuse theefficacy profile.

Long-term efficacy, tolerability and retention

In order to assess the efficacy and safety of a drugwhen administered for extended periods – partic-ularly in chronic diseases, such as epilepsy – long-term follow-up of regulatory studies is required.During open-label follow-up, patients are permit-ted to change the dosing of their medications,providing further useful information. For example,nine cohorts of patients treated with levetiracetamduring its development program were followed fortime periods ranging from 6 months to 5 years (11)(Fig. 2). In addition to demonstrating the long-term efficacy of levetiracetam, these studies alsodemonstrated that 14.4% of patients took fewerconcomitant AEDs at the end of the treatmentperiod than at the start, compared with 6.7% whotook more, and that 5.5% of patients were notreceiving any concomitant AED on their last dayof levetiracetam treatment (11). Audit data areuseful in assessing patient retention in a particulartreatment regimen. For example, a prospectiveanalysis of the outcome of unselected patients whowere prescribed levetiracetam at a regional epilepsyclinic over a 2-year period showed that the cumu-lative probability of remaining on levetiracetamtreatment after 1 year was 74% (12).

Effect on patients with rare seizure types and syndromes

Regulatory studies are unable to assess the efficacyof drugs in patients with rare conditions. Forexample, with the exception of a recent study withlevetiracetam (13), no randomized, controlled trials(RCTs) have examined initial monotherapy oradjunctive therapy in patients with juvenile myo-clonic epilepsy (JME). Current recommendationson the use of lamotrigine, levetiracetam, topira-mate and valproic acid in JME are only availableat recommendation level C and at evidence levelIII, i.e. based either on observational studies or ongeneralizations made on the basis of RCTs con-ducted in patients with other conditions.Information on seizure aggravation from AEDs

in rare conditions can only be obtained when thedrugs are used in clinical practice. For example,clinical evidence in the form of case series or casereports demonstrated that phenytoin, carbamaze-pine, tiagabine, vigabatrin, oxcarbazepine andgabapentin aggravate absence and myoclonic sei-zures and should not be used in JME; andlamotrigine can aggravate severe myoclonic epi-lepsies in infancy and possibly also JME (14–17).

Idiosyncratic side effects, unusual interactions and long-termtoxicity

Long-term toxicity and non-dose-related, idiosyn-cratic side effects with low prevalence may onlymanifest themselves during clinical practice(Table 1). For example, some AEDs are associatedwith weight change (18) or the development ofbone disease (19) over time; Dupuytren’s contrac-tures are a chronic side effect associated withphenobarbital treatment (20); AEDs that altercytochrome P450 enzymes have been shown toaffect the metabolism of sex steroid hormones,leading to reproductive health disorders, sexualdysfunction or disturbance in carbohydrate meta-bolism (21). Several AEDs (phenytoin, phenobar-bital, carbamazepine, oxcarbazepine, lamotrigineand possibly zonisamide) have been associatedwith the so-called anticonvulsant hypersensitivitysyndrome, which is characterized by the triad offever, rash and internal organ involvement andcross-reactivity among the different AEDs (22).Regulatory studies cannot predict unusual inter-

actions, particularly when strict exclusion criteriaprevent certain patient groups, such as pregnantwomen, from participation. For example, there canbe a gradual decline in the concentration-to-doseratio of lamotrigine during pregnancy, which maylead to an increase in seizures or toxic side effectswhen levels rapidly return to baseline following

Figure 2. Long-term follow-up of levetiracetam studies:median percentage reduction from baseline in seizure fre-quency over time, by duration of exposure to levetiracetam.Each line represents seizure frequency data for a cohort ofpatients followed for the entire indicated duration of treat-ment. Reproduced from Ben-Menachem et al. (11) withpermission from Elsevier.

Regulatory studies: what do and don’t they tell?

23

delivery (23) (Fig. 3). Many AEDs have demon-strated adverse events during pregnancy, whichmay result in stillbirth, perinatal mortality, intra-uterine growth retardation, developmental delay orteratogenicity (24–26). Several AEDs are associ-ated with fetal anticonvulsant syndromes, inclu-ding trimethadione (27, 28), phenytoin (29),valproate (30) and carbamazepine (31). Thesesyndromes are all characterized by facial, lip,cardiac and digital anomalies (32), while some areassociated with other abnormalities: fetal trimeth-adione syndrome, for example, is associated withdysmorphism and mental retardation (28), whilefetal valproate syndrome is characterized by anarray of congenital malformations and facialfeatures (Table 2) (30).Registries have been set up in order to monitor

pregnancy outcomes in women exposed to certainAEDs (33). The International Lamotrigine Preg-nancy Registry, for example, aims to assess therisk of all major malformations associated withfirst trimester exposure to lamotrigine monothera-py and additionally provides data on patterns ofspecific congenital malformations associated with

lamotrigine polytherapy (34). The registry has sofar shown that the observed risk of birth defectsassociated with lamotrigine monotherapy is 2.9%,similar to the rate of 2–3% seen in the generalpopulation (34, 35). However, the risk associatedwith lamotrigine polytherapy that includes val-proate is 12.5%, which is consistent with, albeitslightly higher than, published data on congenitalmalformations associated with valproate mono-therapy (34, 36), whereas the risk associated withlamotrigine polytherapy that does not includevalproate is only 2.7% (34). Audit data of thiskind often provide the only means of assessingthe overall performance of a drug in clinicalpractice.

Conclusions

Regulatory studies provide useful, clear-cut infor-mation about a drug’s efficacy, tolerability andease of use. However, these studies are limited asthey have inflexible designs, limited evaluationperiods and strict inclusion and exclusion criteria,resulting in narrowly defined patient populationsthat are not representative of those encountered in

Table 1 Non-dose-related adverse events associated with the newer AEDs

AED Serious adverse events Non-serious adverse events

Gabapentin None Weight gain, peripheral edema, behavioral changes*Lamotrigine Rash (including Stevens-Johnson and toxic

epidermal necrolysis), hypersensitivity reactionsTics* and insomnia

Levetiracetam None Irritability/behavior changeOxcarbazepine Hyponatremia, rash NoneTiagabine Stupor or spike-wave stupor WeaknessTopiramate Renal calculi, open-angle glaucoma, hypohidrosis* Metabolic acidosis, weight loss, language dysfunctionVigabatrin Peripheral visual field defects Psychosis and depressionZonisamide Rash, renal calculi, hypohidrosis* Irritability, photosensitivity, weight loss

Reproduced with permission from French JA et al. Efficacy and tolerability of the new antiepileptic drugs I: treatment of new-onset epilepsy. Neurology 2004;62:1252–60.*Children only or predominantly children.

Figure 3. Lamotrigine concentration-to-dose ratios in 11 ret-rospective pregnancies, expressed as percentage of baseline.Reproduced with permission from de Haan et al. (23).

Table 2 Characteristics of fetal valproate syndrome (30)

Congenital malformations Facial features

Neural tube defects TrigonocephalyCongenital heart disease Tall forehead with bifrontal narrowingCleft lip and palate Epicanthic foldsGenitourinary malformations Infraorbital grooveTracheaomalacia Medial deficiency of eyebrowsRadial ray defects Flat nasal bridgeArachnodactyly/overlapping digits Broad nasal rootAbdominal wall defects Anteverted nostrils

Shallow philtrumLong upper lip with thin vermillion borderThick lower lipSmall down-turned mouth

Ben-Menachem

24

clinical practice. Phase IV studies, long-term open-label follow-up of regulatory studies and audit dataare essential in order to fully define a drug’sperformance in everyday clinical practice, bybacking up evidence with experience and allowinga thorough assessment of benefit and risk.

Acknowledgement

This paper was supported by a grant from UCB.

References

1. Mignot G. Drug trials in epilepsy. BMJ 1996;313:1158.2. Shorvon SD, Lowenthal A, Janz D, Bielen E, Loiseau P, for

the European Levetiracetam Study Group. Multicenterdouble-blind, randomized, placebo-controlled trial oflevetiracetam as add-on therapy in patients with refractorypartial seizures. Epilepsia 2000;41:1179–86.

3. Cereghino JJ, Biton V, Abou-Khalil B, Dreifuss F, Gauer LJ,

Leppik I, the United States Levetiracetam Study Group.Levetiracetam for partial seizures: results of a double-blind, randomized clinical trial. Neurology 2000;55:236–42.

4. Betts T, Waegemans T, Crawford P. A multicentre, double-blind, randomized, parallel group study to evaluate thetolerability and efficacy of two oral doses of levetiracetam,2000 mg daily and 4000 mg daily, without titration inpatients with refractory epilepsy. Seizure 2000;9:80–7.

5. Boon P, Chauvel P, Pohlmann-Eden B, Otoul C, Wroe S.Dose–response effect of levetiracetam 1000 and 2000 mg/day in partial epilepsy. Epilepsy Res 2002;48:77–89.

6. Ben-Menachem E, Falter U, for the European Levetiracetam

Study Group. Efficacy and tolerability of levetiracetam3000 mg/d in patients with refractory partial seizures: amulticenter, double-blind, responder-selected study evalu-ating monotherapy. Epilepsia 2000;41:1276–83.

7. Ben-Menachem E. Preliminary efficacy of levetiracetam inmonotherapy. Epileptic Disord 2003;5(Suppl. 1):S51–5.

8. Steinhoff B, Van Paesschen W, Sadzot B, Maubrey MC,

Genton P. SKATE trial interim analysis: confirmation ofsafety and efficacy of levetiracetam (Keppra�) as adjunc-tive therapy in patients with refractory epilepsy. Posterpresented at the 25th International Epilepsy Congress,Lisbon, Portugal, 2003; abstract number P463. Availablefrom: http://www.epilepsycongress.org/abs_index.asp. Ac-cessed January 24, 2005.

9. Morrell MJ, Leppik I, French J, Ferrendelli J, Han J, Magnus

L. The KEEPER trial: levetiracetam adjunctive treatmentof partial-onset seizures in an open-label community-basedstudy. Epilepsy Res 2003;54:153–61.

10. Ferrendelli JA, French J, Leppik I et al. Use of levetirace-tam in a population of patients aged 65 years and older: asubset analysis of the KEEPER trial. Epilepsy Behav2003;4:702–9.

11. Ben-Menachem E, Edrich P, Van Vleymen B, Sander JW,

Schmidt B. Evidence for sustained efficacy of levetiracetamas add-on epilepsy therapy. Epilepsy Res 2003;53:57–64.

12. Nicolson A, Lewis SA, Smith DF. A prospective analysis ofthe outcome of levetiracetam in clinical practice. Neurol-ogy 2004;63:568–70.

13. Verdru P, Wajgt A, Schiemann Delgado J. Efficacy andsafety of levetiracetam 3000 mg/d as adjunctive treatmentin adolescents and adults suffering from idiopathic

generalized epilepsy with myoclonic seizures. Epilepsia2005;46(Suppl. 6):56–7.

14. Carrazana EJ, Wheeler SD. Exacerbation of juvenilemyoclonic epilepsy with lamotrigine. Neurology 2001;56:1424–5.

15. Guerrini R, Belmonte A, Genton P. Antiepileptic drug-in-duced worsening of seizures in children. Epilepsia1998;39(Suppl. 3):S2–10.

16. Genton P. When antiepileptic drugs aggravate epilepsy.Brain Dev 2000;22:75–80.

17. Bourgeois BF. Reducing overtreatment. Epilepsy Res2002;52:53–60.

18. Biton V. Effect of antiepileptic drugs on bodyweight:overview and clinical implications for the treatment ofepilepsy. CNS Drugs 2003;17:781–91.

19. Pack AM. The association between antiepileptic drugs andbone disease. Epilepsy Curr 2003;3:91–5.

20. Critchley EM, Vakil SD, Hayward HW, Owen VM. Dup-uytren’s disease in epilepsy: result of prolonged adminis-tration of anticonvulsants. J Neurol Neurosurg Psychiatry1976;39:498–503.

21. Morrell MJ, Flynn KL, Seale CG et al. Reproductivedysfunction in women with epilepsy: antiepileptic drugeffects on sex-steroid hormones. CNS Spectr 2001;6:771–86.

22. Knowles SR, Shapiro LE, Shear NH. Anticonvulsanthypersensitivity syndrome: incidence, prevention andmanagement. Drug Saf 1999;21:489–501.

23. de Haan GJ, Edelbroek P, Segers J et al. Gestation-inducedchanges in lamotrigine pharmacokinetics: a monotherapystudy. Neurology 2004;63:571–3.

24. Holmes LB, Harvey EA, Coull BA et al. The teratogenicityof anticonvulsant drugs. N Engl J Med 2001;344:1132–8.

25. Tettenborn B, Genton P, Polson D. Epilepsy and women’sissues: an update. Epileptic Disord 2002;4(Suppl. 2):S23–31.

26. Yerby MS. Special considerations for women with epilepsy.Pharmacotherapy 2000;20:159S–170S.

27. German J, Ehlers KH, Kowal A, De George FV, Engle MA,

Passarge E. Possible teratogenicity of trimethadione andparamethadione. Lancet 1970;2:261–2.

28. Zackai EH, Mellman WJ, Neiderer B, Hanson JW. The fetaltrimethadione syndrome. J Pediatr 1975;87:280–4.

29. Hanson JW, Smith DW. The fetal hydantoin syndrome.J Pediatr 1975;87:285–90.

30. DiLiberti JH, Farndon PA, Dennis NR, Curry CJ. The fetalvalproate syndrome. Am J Med Genet 1984;19:473–81.

31. Jones KL, Lacro RV, Johnson KA, Adams J. Pattern ofmalformations in the children of women treated withcarbamazepine during pregnancy. N Engl J Med1989;320:1661–6.

32. Pennell PB. Pregnancy in the woman with epilepsy: mater-nal and fetal outcomes. Semin Neurol 2002;22:299–308.

33. Beghi E, Annegers JF, Collaborative Group for the Preg-

nancy Registries in Epilepsy. Pregnancy registries in epi-lepsy. Epilepsia 2001;42:1422–5.

34. Cunnington MC. The International Lamotrigine PregnancyRegistry update for the epilepsy foundation. Epilepsia2004;45:1468.

35. Honein MA, Paulozzi LJ, Cragan JD, Correa A. Evaluationof selected characteristics of pregnancy drug registries.Teratology 1999;60:356–64.

36. Samren EB, van Duijn CM, Koch S et al. Maternal use ofantiepileptic drugs and the risk of major congenital mal-formations: a joint European prospective study of humanteratogenesis associated with maternal epilepsy. Epilepsia1997;38:981–90.

Regulatory studies: what do and don’t they tell?

25