OR WAIT null SECS
Steven Denham, director of biostatistics at MPI Research, discusses the impact the Standard for Exchange of Nonclinical Data may have on the pharmaceutical industry.
The Clinical Data Interchange Standards Consortium (CDISC), a nonprofit standards development organization, developed the Standard for Exchange of Nonclinical Data (SEND) as a method to standardize the way pharmaceutical companies and contract research organizations (CROs) submit electronic data to FDA. Work on the standard began in 2002, with the first really complete Implementation Guide (SEND-IG 3.0) issued in 2011. SEND is a nonclinical version of the Study Data Tabulation Model (SDTM), the standard format for electronic submission of clinical data to FDA. SEND and SDTM are expected to speed up the review process for drug applications by “developing electronic tools to analyze and visualize these submissions, and building data warehouses to rapid query data across drugs, companies, and clinical and nonclinical disciplines” (1).
In December 2014, FDA released guidance on Providing Regulatory Submissions in Electronic Format--Standardized Study Data, citing the agency’s updated policies for electronic submission of datasets (2). In the guidance, the agency sites SEND and SDTM as the preferred formats for submitting electronic data for review. Although SEND datasets are not currently required for submission, by Dec. 17, 2016 SEND will become a mandatory part of submissions for relevant new drug applications (NDAs), biologics license applications (BLAs), and abbreviated new drug applications (ANDAs). The datasets will be required for submission of relevant investigational new drugs (INDs) by Dec. 17, 2017.
In upcoming months, pharmaceutical companies will have to work to format data according to SEND standards. CDISC released the SEND Implementation Guide (SEND IG) to assist companies in the transition, and some CROs have created SEND dataset resources for clients. Within the next year, some pharmaceutical companies may have to restructure pre-established data systems in order to comply with SEND. If companies fail to comply, they could face refusal to file from FDA, delaying application review and approval processes.
Steven Denham, PhD, director, biostatistics and information sciences, at MPI Research sees opportunities and obstacles ahead for companies with the implementation of SEND. Pharmaceutical Technology sat down with Denham to discuss the overall impact SEND will have on the pharmaceutical industry.
SEND guidelines and regulationsPharmTech: Can you give a brief overview of SEND?
Denham: The idea is to provide an electronic standard for capturing individual data from nonclinical studies. The CDISC standards require the production of datasets that includes both metadata about how the study is conducted, and the actual data itself. The full SEND package also contains DEFINE files, which are part of any CDISC compliant package. They contain no data themselves, but do contain information defining all variables in the other domains. The package also includes the RELREC domain that relates records in one domain file to records in another domain file, and a study data reviewers’ guide. So, I think those are the parts that all fit into it. There is more than just data in a SEND package.
PharmTech: What was the thinking behind SEND?
Denham: I know FDA is trying to move away from paper that is creating a physical storage problem. Plus, they have experience on the clinical side in getting their individual data in electronic format. The SEND initiative is really an outgrowth of those two thoughts. Eventually, it should improve the review process, not only for review time, but [also] for the safety aspects in that there would be a data warehouse at the agency that can be searched across studies and across sponsors. Therefore, there is this push to make SEND a requirement.
PharmTech: What are some of the issues that FDA encountered that led to the creation of SEND?
Denham: First, how do you store truckloads of data and be able to access it in a finite review time? The second issue is that none of these studies are large enough to outline all of the safety issues that you might find going into human trials. An agency review team might say, ‘We have three rat studies, two dog studies and a monkey study for this compound. Let’s check our data warehouse for studies that involved a similar active ingredient.’ Those studies could well have been submitted by different sponsors. The end result is that they can recommend a more focused approach to safety observations in clinical trials.
SEND’s impact on the pharma industryPharmTech: What is the biggest way that SEND will impact the pharmaceutical/biopharmaceutical industry in the future?
Denham: In the future, SEND will enable FDA to speed up their review. Also, SEND will make it easier on sponsors to pull together parts of studies or even multiple studies conducted in house, by [contract research organizations] CROs, and by a second CRO, all into one collection of datasets that are formatted consistently.
PharmTech: What are some of the ways the SEND guidance will impact the way CROs communicate with clients?
Denham: The SEND guidance will impact communication with sponsors. Currently, reports are customizable for every sponsor. Therefore, we’ve gone to a lot of trouble, and I know other CROs have as well, to make our reports customizable for each sponsor.
SEND does exactly the opposite. Basically it states that everybody will receive a standard representation of the individual data in the same format. There are strict rules for how it’s going to be presented, how things are defined, what terms are used.
Another thing that is going to change is a SEND data package is designed to reflect the individual data in the final report. Consequently, a full and complete SEND data package can’t be assembled until the study is finalized.
Now, a large number of sponsors are setting up all of their in-house databases to input material in SEND format. However, they also like to have interim data available as a study is going on. That’s fine, but it means pulling what has been kind of easy into a very defined format before sending that on.
This has triggered some concern from the agency in that SEND datasets are being used to monitor outcomes of a trial. This would cause issues after finalization of the data, especially if you decide to change the study design or add on any components, as SEND datasets are to be organized at the completion of the studies.
However, the concern from the agency arises using interim data to make decisions regarding the conduct of the study, and also submitting that same data to the agency. Then it becomes a Part 58 issue, and subject to study director single point of control and [quality assurance] QA unit audit, which requires time to complete, and rather defeats the purpose of interim data for study monitoring.
Still, this is all in a formative stage. Historically, however, when the agency has said something may be required (QA audit of SEND formatted interim data sets) in the future, it almost always does turn out to be required.
Currently, SEND data packages do not require a full QA audit, but if they’re going to be used for interims, it looks like perhaps they will [be required]. So, I guess those are the biggest challenges going on.
CROs and SEND compliancePharmTech: Is this something all CROs will have to comply with in upcoming years?
Denham: In order for a CRO to remain successful they will have to be able to produce SEND data packages, just as they now have to produce eCTDs (electronic common technical documents). As you may know, if you are a CRO and don’t produce eCTDs, you are out of business very quickly. It’s going to be the same way with SEND data packages.
PharmTech: Will CROs have to restructure in order to comply with this guidance?
Denham: In my opinion, CROs that have a strong clinical presence are equipped with an entire professional group of data mangers, which includes data entry staff, report staff, and statisticians. The same will be required for non-clinical data. I think that’s probably the biggest change, and it’s one of the hardest for managers of CROs to accept.
For those that don’t operate in the clinical space, they are accustomed to data input, data output, and report production. With SEND, many of the connections between observations have to be brought in electronically as much as they are in the results and discussion section of the final report. For example, a sponsor was thinking that they would have their in-house, board certified, toxicologists creating the SEND data sets. This would be a wasteful use of resources, because this is a data management role and pattern recognition type job not a biological or scientific job function.
PharmTech: What is some other information you feel is important for companies to know about SEND?
Denham: Everyone sees the FDA guidance and thinks that SEND data packages are going to be required, but it’s only for a limited number of study types at this point. The SEND IG does not yet model safety pharmacology study, developmental and reproductive toxicology studies, gene-tox studies, or dermal studies.
There are a lot of sponsors [that may think] that SEND data packages are required for every study, we have seen sponsors panicking over the development of SEND datasets for all of their studies being conducted. The important thing to do is to be sure about the types of studies you are conducting and deciding whether SEND datasets are required.
1. T Anzai et al. Journal of Toxicologic Pathology online, DOI: 10.1293/tox.2015-0007 (April 1, 2015).
2. FDA, Guidance for Industry, Providing Regulatory Submissions In Electronic Format--Standardized Study Data (Silver Spring, MD, Dec. 17, 2014).
Article DetailsPartnerships in Outsourcing Supplement
Vol. 40, No. 13
When referring to this article, please cite it as C. Hroncich, “The Impact of SEND on the Pharmaceutical Industry," Pharmaceutical Technology Partnerships in Outsourcing Supplement 40 (13) 2016.