Discontinuation and Nonpublication of Randomized Clinical Trials Conducted in Children

Drs. Florence Bourgeois and Natalia Pica – pediatricians at Harvard Medical School and Boston Children’s Hospital – recently published a manuscript in Pediatrics (Pediatrics 2016; 138: 1-9) investigating interventional clinical trials in children conducted between 2008 and 2010. Their paper revealed two “troubling” findings. First, that 104 of the 559 trials started (19%) were discontinued before they were completed. Second, of the 455 completed trials, 136 (30%) were not published. As Bourgeois and Pita note, there is an ethical imperative to ensure that the results of clinical trials are reported and that the risks taken by the children who participated in the research benefits all children. Determinants of study discontinuation and publication are complex. Discontinuation was less common among industry funded than among investigator-initiated studies and failure to publish was more common among industry-funded studies. Many unpublished studies were funded by classical academic sources and conducted in leading academic centers. Clearly other factors are at play. A recent commentary on NPR Health News reviewing this study suggested that this occurs because investigators “don’t take the effort to publish and share their results.” Having published many clinical trials, we found this assessment is unfair to the researchers and fails to identify likely causes. Several practical issues make publication of pediatric clinical trials difficult. Pediatric trials are often small, especially when compared to many adult studies, which detracts from their perceived scientific and societal impact and negatively affects publication decisions. Trials involving children are more likely to be published in pediatric journals as opposed to more widely read mainstream biomedical journals. This is especially true for studies reporting negative results, a factor known to be associated with editorial bias. Pediatric trials are often conducted long after adult approval when “off-label” pediatric treatment is frequent which reduces parent, practitioner and investigator enthusiasm contributing to small studies. Pediatric studies may be conducted to satisfy a regulatory requirement rather than to generate new knowledge to improve pediatric therapy reducing both investigator and publisher enthusiasm, again contributing to small and hard to publish studies. Finally, clinical trials initiated after September 27, 2007 that meet the FDAAA 801 definition of an "applicable clinical trial, conducted at one or more sites in the U.S. or those supported by funding from the U.S. National Institutes of Health must register at ClinicalTrials.gov. While this mechanism does make information from specific trials generally accessible, it is not a substitute for the proven, time-honored peer review publication process. In conclusion, child health researchers are by and large hard working, ethical and highly motivated professionals who are committed to improving the health and welfare of children through the process of discovery, which includes conducting and publishing research to the highest standard. The implication that the publication gap is due to lack of effort is, in our view, neither accurate nor fair.

Dear Editors, August 23, 2016

It was with great interest that we read the results of Doctors Pica and Bourgeois’ August 4th, 2016 article on “Discontinuation and Nonpublication of Randomized Clinical Trials Conducted in Children”. As they noted, enrollment of children into discontinued and/or unpublished trials is concerning from both scientific and ethical perspectives. We agree information from these trials should be publicly available and critically reviewed. We would, however, like to note that in addition to peer reviewed publications, there is another source of information on pediatric trials that ensures the interpretation of the data from the trials “are rigorously scrutinized and ensures appropriate interpretation of the results.”

The authors refer to the Pediatric legislation that has greatly facilitated FDA’s ability to obtain pediatric trials for products that are being developed for adults by pharmaceutical companies.
This same legislation also requires the FDA to post its reviews of the Medical, Pharmacology and Statistical data submitted for pediatric trials. These reviews are publicly available on FDA’s website at http://www.fda.gov/ScienceResearch/SpecialTopics/PediatricTherapeuticsRe... and provide a rich source of information on pediatric trials. FDA has now reviewed pediatric studies for over 630 products and there is an enormous amount of information to be learned from both the failed and successful trials. Under the Best Pharmaceutical for Children legislation, even ”failed” trials may result in the product receiving additional marketing exclusivity, IF they have performed the trials in the manner requested by FDA. Ongoing analysis of these trials has revealed errors in our assumptions about the similarity and differences between adults and children in both the pathophysiology of some diseases and the responses to therapies administered as treatment. (1,2,3)

We agree with the authors that all data obtained from children who are participating in a trial should be publicly available in order to advance our knowledge, even if the trial was not completed or failed to demonstrate the desire or anticipated response. These data are actually sometimes the most informative information we obtain. They challenge our preconceived ideas or assumptions and advance our knowledge about possible differences between the pediatric and adult manifestations of disease and responses to therapy.

Sincerely,
Dianne Murphy, MD*
Director, Office of Pediatric Therapeutics, FDA
[email protected]
301-796-8651 10903 New Hampshire Ave., WO32-5150
Silver Spring, MD 20993
*The views and opinions expressed are those of the author and do not reflect official FDA opinion or policy
References:
1) Failed Pediatric Trials of Acute Migraine Treatments: Lessons Learned From a Systematic Review of Trial Data Submitted to the Food and Drug Administration (FDA). Haihao Sun, MD, PhD, Eric Bastings, MD, Jean Temeck, MD, P. Brian Smith, MD MPH MHS, Angela Men, MD, Veneeta Tandon, PhD, Dianne Murphy, MD; William Rodriguez, MD, PhD
JAMA Pediatrics March 2013:167 (3) 243 -249
http://archpedi.jamanetwork.com/Mobile/article.aspx?articleid=1558561
2) Safety and Transparency of Pediatric Drug Trials
Daniel K. Benjamin Jr, MD, PhD, MPH; P. Brian Smith, MD, MHS; M. Jessica M. Sun, MD; M. Dianne Murphy, MD; Debbie Avant, RPh; Lisa Mathis, MD; William Rodriguez, MD; Robert M. Califf, MD; Jennifer S. Li, MD, MHS http://www.ncbi.nlm.nih.gov/pubmed/19996043
Arch Pediatr Adolesc Med. December 2009;163(12):1080-1086 **
3) Pediatric Antihypertensive Trial Failures: Analysis of Endpoints and Dose Range
Daniel K. Benjamin, Jr., MD, PhD, MPH; P. Brian Smith, MD; Pravin Jadhav, PhD; Jogarao V. Gobburu, PhD; M. Dianne Murphy, MD; Victor Hasselblad, PhD; Carissa Baker-Smith, MD; Robert M. Califf, MD; and Jennifer S. Li, MD, MHS
Hypertension 2008; 51; 834-840, March 10, 2008
http://www.ncbi.nlm.nih.gov/pubmed/18332283