Accepted Manuscript Extrapolation will never replace randomized clinical trials Perrine Janiaud, Catherine Cornu, Behrouz Kassai PII:
S0895-4356(15)00332-7
DOI:
10.1016/j.jclinepi.2015.06.019
Reference:
JCE 8936
To appear in:
Journal of Clinical Epidemiology
Received Date: 6 May 2015 Revised Date:
29 May 2015
Accepted Date: 17 June 2015
Please cite this article as: Janiaud P, Cornu C, Kassai B, Extrapolation will never replace randomized clinical trials, Journal of Clinical Epidemiology (2015), doi: 10.1016/j.jclinepi.2015.06.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Extrapolation will never replace randomized clinical trials Perrine Janiaud c, Catherine Cornua b c,Behrouz Kassaia b c a
Hôpital Louis Pradel, EPICIME-Centre d'Investigation Clinique, CIC201/UMR5558, 28, Avenue du Doyen Lépine, 69677 cedex, Bron, France;
INSERM
b
RI PT
CHU Lyon, Service de Pharmacologie Clinique, 8 rue Guillaume Paradin, BP8071, 69376 cedex 08, Lyon, France;
c
University of Lyon 1, UMR 5558, CRNS Lyon, 8 rue Guillaume Paradin, BP8071, 69376 cedex 08, Lyon, France Abstract:
SC
Current regulatory guidelines suggest that randomized controlled trials for children become mandatory when there are uncertainties on PK-PD data. Our results show, however,
M AN U
that the similarities between efficacies of medicines evaluated both in adults and children are commonly unclear, sometimes conflicting or differ quantitatively. We suggest that the extrapolation process for drug development and clinical practice should be based on a prior synthesis of all available evidence in adults and children and include at least one randomized controlled trial in children. Response to commentary by R. Oostenbrink and S.N. de Wildt (1) on our article (2).
TE D
Keywords: Extrapolation; Treatment benefit; Pediatrics; Adults; Randomized controlled trials Response to commentary:
We thank R Oostenbring and S de Wildt (1) for their comment. In our article (2), we do not challenge the current paradigm that “drugs need to be studied in children”. We rather
EP
suggest that the extrapolation framework could be improved both for drug development and for routine practice by using all available evidence. Currently, European Medicines Agency
AC C
(EMA) and Food and Drug Administration (FDA) guidelines suggest that when there are robust pharmacokinetic (PK) and pharmacodynamic (PD) data, and the disease progression is similar between adults and children, a full extrapolation can be conducted after conducting a safety study and/or a PK study (3, 4). When there are uncertainties on PK-PD, data on efficacy are required and randomized controlled trials become mandatory. The extrapolation process is not limited to drug development, because pediatricians routinely extrapolate results of adult studies for treating children without a clear framework. Our contribution shows that we lack data on the dose–effect relationship for different age groups and that the similarities between efficacies of medicines evaluated both in adults and children are commonly unclear, sometimes conflicting or differ quantitatively. With only 36 out of 74 treatments effective in both populations, our study illustrates that even for the same disease 1
ACCEPTED MANUSCRIPT and same medicines the clinical efficacy from adults to children cannot be extrapolated systematically. We believe that the extrapolation process should include at least one randomized controlled trial (RCT) in children with a prior mandatory evaluation of all available evidence in adults and children. Figure 1 demonstrates for ondansetron in preventing postoperative
RI PT
nausea and vomiting, how adult trials could be used before launching studies in children. We constructed an effect model (green line) which is simply the regression line relating the risk in the treated and in the untreated adult population based on the results of RCTs (5). Such a model allows the prediction of the potential treatment benefit in children before designing a new trial. It improves the calibration of pediatric trial according to the baseline risk, the
SC
identification of children who would potentially benefit more from the treatment, and importantly avoids unnecessary trials in children. In this example, 27 trials have been
M AN U
performed in children! Nineteen were performed after 1995 including 1,662 children, when the treatment benefit has been definitely shown.
Finally, our study shows that the lack of standardized reports limits the feasibility of and the access to a comprehensive synthesis of all available evidence for guiding drug development and clinical practice (6, 7).
TE D
References:
AC C
EP
1. Oostenbrink R, de Wildt S. Drug trials: Kids are no little adults and not all kids are the same. J Clin Epid [Internet]. 2015; 15-133. 2. Janiaud P, Lajoinie L, Fleur C-A, Cornu C, Cochat P, Cucherat M, et al. Similarity of the treatment benefit between children and adults: Results of a metaepidemiologicalstudy. J Clin Epid [Internet]. 2015; 13-809-R2. 3. European Medicines Agency. Concept paper on extrapolation of efficacy and safety in medicine development EMA/129698/2012 Human Medicines Development and Evaluation. 2012. 4. Dunne J, Rodriguez WJ, Murphy MD, Beasley BN, Burckart GJ, Filie JD, et al. Extrapolation of adult data and other data in pediatric drug-development programs. Pediatrics. 2011 Nov;128(5):e1242-9. doi: 10.542/peds.2010-3487. Epub 2011 Oct 24. 5. Boissel JP, Cucherat M, Nony P, Chabaud S, Gueyffier F, Wright JM, et al. New insights on the relation between untreated and treated outcomes for a given therapy effect model is not necessarily linear. J Clin Epidemiol. 2008 Mar;61(3):301-7. doi: 10.1016/j.jclinepi.2007.07.007. Epub Nov 28. 6. Clarke M, Brice A, Chalmers I. Accumulating research: a systematic account of how cumulative meta-analyses would have provided knowledge, improved health, reduced harm and saved resources. PLoS One. 2014 Jul 28;9(7):e102670. doi: 10.1371/journal.pone.0102670. eCollection 2014. 7. Glasziou P. Paul Glasziou: Six proposals for evidence based medicine’s future 2015 [cited 2015 14th April 2015]. Available from: http://blogs.bmj.com/bmj/2015/03/27/paulglasziou-six-proposals-for-evidence-based-medicines-future/.
2
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1: The effect model for ondansetron to prevent postoperative nausea and vomiting using adults RCTs is illustrated by the green line. Rc corresponds to the risk of presenting the event in the untreated patients or the control group and Rt the risk in the treated patients. The dashed line Rt=Rc is the no effect line, when Rt < Rc, below the dashed line, treatment is effective and when Rt>Rc, above the dashed line, treatment is harmful. The full line is a linear regression that fit the data in adults. The effect model shows a beneficial effect for ondansetron in the prevention of postoperative nausea and vomiting in adults. The resulting equation of the adult RCTs (green squares) (Rt=0.63Rc-0.03) can then be used to predict the efficacy in children RCTs (red circles) using the baseline risk. The children trials have also been modeled (red line) on the plot confirming that the effect model in adults gives a good prediction of the treatment benefit in children.
3
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0895-4356(15)00332-7
DOI:
10.1016/j.jclinepi.2015.06.019
Reference:
JCE 8936
To appear in:
Journal of Clinical Epidemiology
Received Date: 6 May 2015 Revised Date:
29 May 2015
Accepted Date: 17 June 2015
Please cite this article as: Janiaud P, Cornu C, Kassai B, Extrapolation will never replace randomized clinical trials, Journal of Clinical Epidemiology (2015), doi: 10.1016/j.jclinepi.2015.06.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Extrapolation will never replace randomized clinical trials Perrine Janiaud c, Catherine Cornua b c,Behrouz Kassaia b c a
Hôpital Louis Pradel, EPICIME-Centre d'Investigation Clinique, CIC201/UMR5558, 28, Avenue du Doyen Lépine, 69677 cedex, Bron, France;
INSERM
b
RI PT
CHU Lyon, Service de Pharmacologie Clinique, 8 rue Guillaume Paradin, BP8071, 69376 cedex 08, Lyon, France;
c
University of Lyon 1, UMR 5558, CRNS Lyon, 8 rue Guillaume Paradin, BP8071, 69376 cedex 08, Lyon, France Abstract:
SC
Current regulatory guidelines suggest that randomized controlled trials for children become mandatory when there are uncertainties on PK-PD data. Our results show, however,
M AN U
that the similarities between efficacies of medicines evaluated both in adults and children are commonly unclear, sometimes conflicting or differ quantitatively. We suggest that the extrapolation process for drug development and clinical practice should be based on a prior synthesis of all available evidence in adults and children and include at least one randomized controlled trial in children. Response to commentary by R. Oostenbrink and S.N. de Wildt (1) on our article (2).
TE D
Keywords: Extrapolation; Treatment benefit; Pediatrics; Adults; Randomized controlled trials Response to commentary:
We thank R Oostenbring and S de Wildt (1) for their comment. In our article (2), we do not challenge the current paradigm that “drugs need to be studied in children”. We rather
EP
suggest that the extrapolation framework could be improved both for drug development and for routine practice by using all available evidence. Currently, European Medicines Agency
AC C
(EMA) and Food and Drug Administration (FDA) guidelines suggest that when there are robust pharmacokinetic (PK) and pharmacodynamic (PD) data, and the disease progression is similar between adults and children, a full extrapolation can be conducted after conducting a safety study and/or a PK study (3, 4). When there are uncertainties on PK-PD, data on efficacy are required and randomized controlled trials become mandatory. The extrapolation process is not limited to drug development, because pediatricians routinely extrapolate results of adult studies for treating children without a clear framework. Our contribution shows that we lack data on the dose–effect relationship for different age groups and that the similarities between efficacies of medicines evaluated both in adults and children are commonly unclear, sometimes conflicting or differ quantitatively. With only 36 out of 74 treatments effective in both populations, our study illustrates that even for the same disease 1
ACCEPTED MANUSCRIPT and same medicines the clinical efficacy from adults to children cannot be extrapolated systematically. We believe that the extrapolation process should include at least one randomized controlled trial (RCT) in children with a prior mandatory evaluation of all available evidence in adults and children. Figure 1 demonstrates for ondansetron in preventing postoperative
RI PT
nausea and vomiting, how adult trials could be used before launching studies in children. We constructed an effect model (green line) which is simply the regression line relating the risk in the treated and in the untreated adult population based on the results of RCTs (5). Such a model allows the prediction of the potential treatment benefit in children before designing a new trial. It improves the calibration of pediatric trial according to the baseline risk, the
SC
identification of children who would potentially benefit more from the treatment, and importantly avoids unnecessary trials in children. In this example, 27 trials have been
M AN U
performed in children! Nineteen were performed after 1995 including 1,662 children, when the treatment benefit has been definitely shown.
Finally, our study shows that the lack of standardized reports limits the feasibility of and the access to a comprehensive synthesis of all available evidence for guiding drug development and clinical practice (6, 7).
TE D
References:
AC C
EP
1. Oostenbrink R, de Wildt S. Drug trials: Kids are no little adults and not all kids are the same. J Clin Epid [Internet]. 2015; 15-133. 2. Janiaud P, Lajoinie L, Fleur C-A, Cornu C, Cochat P, Cucherat M, et al. Similarity of the treatment benefit between children and adults: Results of a metaepidemiologicalstudy. J Clin Epid [Internet]. 2015; 13-809-R2. 3. European Medicines Agency. Concept paper on extrapolation of efficacy and safety in medicine development EMA/129698/2012 Human Medicines Development and Evaluation. 2012. 4. Dunne J, Rodriguez WJ, Murphy MD, Beasley BN, Burckart GJ, Filie JD, et al. Extrapolation of adult data and other data in pediatric drug-development programs. Pediatrics. 2011 Nov;128(5):e1242-9. doi: 10.542/peds.2010-3487. Epub 2011 Oct 24. 5. Boissel JP, Cucherat M, Nony P, Chabaud S, Gueyffier F, Wright JM, et al. New insights on the relation between untreated and treated outcomes for a given therapy effect model is not necessarily linear. J Clin Epidemiol. 2008 Mar;61(3):301-7. doi: 10.1016/j.jclinepi.2007.07.007. Epub Nov 28. 6. Clarke M, Brice A, Chalmers I. Accumulating research: a systematic account of how cumulative meta-analyses would have provided knowledge, improved health, reduced harm and saved resources. PLoS One. 2014 Jul 28;9(7):e102670. doi: 10.1371/journal.pone.0102670. eCollection 2014. 7. Glasziou P. Paul Glasziou: Six proposals for evidence based medicine’s future 2015 [cited 2015 14th April 2015]. Available from: http://blogs.bmj.com/bmj/2015/03/27/paulglasziou-six-proposals-for-evidence-based-medicines-future/.
2
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 1: The effect model for ondansetron to prevent postoperative nausea and vomiting using adults RCTs is illustrated by the green line. Rc corresponds to the risk of presenting the event in the untreated patients or the control group and Rt the risk in the treated patients. The dashed line Rt=Rc is the no effect line, when Rt < Rc, below the dashed line, treatment is effective and when Rt>Rc, above the dashed line, treatment is harmful. The full line is a linear regression that fit the data in adults. The effect model shows a beneficial effect for ondansetron in the prevention of postoperative nausea and vomiting in adults. The resulting equation of the adult RCTs (green squares) (Rt=0.63Rc-0.03) can then be used to predict the efficacy in children RCTs (red circles) using the baseline risk. The children trials have also been modeled (red line) on the plot confirming that the effect model in adults gives a good prediction of the treatment benefit in children.
3
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
