Far from being simply a summary of individual results, a
comprehensive and detailed summary allows companies to make informed
decisions, explains Michael Whitworth at Quanticate
Producing comprehensive integrated summaries of safety and efficacy is a critical stage of the submission life cycle. These need to be designed and planned carefully in advance to ensure informed decision-making and effectiveness at the regulatory interface. A focus on the approval and whole lifecycle of the product, and not just the submission, will influence the quality and direction of the content. Traceability is key in all respects of the creation of information, from data that leads to knowledgeable decisions and the ultimate wisdom that forms the label of a product.
Guidelines from agencies on the preparation of integrated summaries
for regulatory submissions are often lacking in terms of providing
specific direction on content. What is important to sponsors at the
point of submission can vary, but it is crucial to allocate time and
set a budget for thorough integrated summaries (Integrated Summary of
Safety (ISS) and Integrated Summary of Efficacy (ISE)) that will be
certain of meeting the regulator’s expectations.
An integrated summary should not just be considered a summary of the
details of individual study results held in the Clinical Study Reports
(CSR), but rather, is a comprehensive and in-depth analysis of
aggregated results used to make informed decisions. This analysis
involves a synthesis of the results of individual studies, in an
appropriate manner, to provide evidence of the safety and effectiveness
of the drug. The integrated summary goes beyond the level of a summary,
detailing pooled analyses and discussing them in detail.
EFFECTIVE PLANNING FOR INTEGRATED SUMMARIES
Approvability of a product depends, of course, on the data that is
selected for collection during the planning of clinical trials.
However, considerable resources, time and money can be saved by
planning carefully how the summaries are structured to help make
effective decisions and deliver clear messages supporting the target
product claims. Consideration needs to be made as to where and when the
product will be submitted. Will this strategy for creating the Common
Technical Document (CTD) be discussed with authorities and will
briefing documents be needed? What studies or populations are needed to
fully support the proposed label? Considerations for maximising the
re-use of information are important at this stage. If you are planning
several submissions, it is important to manage the traceability of the
information from the collected raw data through to the different
derived summaries. Managing the consistent derivation of values also
needs the effective management of metadata for later use in submission
and regulatory defence.
Analysis planning of these summaries may be started alongside the
preparation of the Phase II studies to ensure appropriate endpoints,
time points and patient populations are being considered. Visualisation
of the summaries in terms of creation of output templates is defined at
this early stage. These will undergo multiple reviews and updates as
the knowledge of the drug increases and as the process becomes more
focused, with data becoming available from early phase studies. Close
communication of the relevant programmer, statistician, medical writer
and physician is crucial to ensuring that reviews are performed at
sufficiently early time points to allow for setup and reporting in a
timely fashion. Forming an integrated, well-communicated and close-knit
team during the development/review cycle of the summaries will ensure
clearer understanding and high quality documents delivered to tight
timelines. The close partnership of programmer and statistician
provides an integrated approach to a highly technical piece of the CTD
with as much detail as possible. Standard summary templates will
provide a guideline initially, but will likely require amendments for
these summaries.
The planning in terms of how data, information and knowledge are
stored is also important to consider at an early stage. During a
product’s lifecycle the data points collected in clinical studies can
be used in endless ways for information creation and decision points.
Building the right information store and simple traceability will mean
that you will be able to easily see the basis upon which data decisions
were made and, more importantly, effectively respond to questions
during regulatory defence. Including the planning upfront and using
effective tools to manage the traceability will have enormous benefits
to your products label in the long term. Every piece of data,
information and knowledge that is created in the product’s lifecycle
supports the statements made in the label for the safety, efficacy and
populations defined. These pieces of intellectual property are the
crown jewels of your organisation and need to be planned carefully,
documented through metadata and shared with consideration.
CREATION OF INTEGRATED SUMMARIES
It is normally a dash to the finish line when the integrated
summaries are created. The wait for the data to be delivered for the
final clinical study reports can get frustrating as data is being
cleaned and the individual study team process the data. Hopefully, by
this time your programs are ready to produce all the summary tables
that were defined in the templates, and they have been tested and
validated to run like a validated application that is robust enough to
tackle every eventuality of data. Your project teams should have
reviewed the tables using blinded and unblinded data to make sure that
they have all the graphs and summary tables required to best interpret
the information.
Depending on the company/supplier setup, coordination of the review
cycle will generally be done within the biometrics team. Getting the
input of medical writers and physicians before the unblinding of the
Phase III studies is vital but difficult to achieve. Planning review
meetings early in people’s calendars and making sure the relevant
personnel have performed a quality review prior to the meeting to
enable a productive review can be critical in saving on re-engineering
of the project’s critical path.
Delivering an electronic European submission
|
Project scope
A CRO was contracted to produce reports, including clinical study
reports (CSRs), for two placebo controlled trials, along with two
further CSRs for a sponsor’s open-label extension studies. All four
studies were reported from disparate databases, required statistical
analysis plans (SAPs) and CSRs, as well as an ISS and ISE for Module
2.7 and input into Module 2.5. The project started in December 2008 and
was delivered on time in September 2009.
Execution
The CRO introduced a standard naming convention across all databases.
The SAPs and mock-up tables were all produced in a standard fashion,
taking care not to introduce unnecessary inconsistencies. A significant
amount of time was spent standardising the data into similar formats of
derived datasets. This was time well spent, as once the data were in
standard format, it was possible to write a single programme to produce
output for the four studies.
Tight timelines and a heavy workload made it necessary to create teams
of programmers and statisticians who worked on specific endpoints
across all studies to maximise re-use of code, rather than separate
teams working on individual studies.
Much of the effort was focused on writing detailed analysis plans to
ensure that rework post database lock was minimal; all definitions were
clearly defined and there were a significant amount of programmer notes
to reduce the risk of misunderstanding.
All studies had two dry runs to iron out any issues. It was important
to dedicate time to making sure that sufficient peer review was
undertaken to ensure consistency between endpoints programmed by
different teams.
The four databases were frozen between late April and early May. All
tables, listings, figures and statistical reports were finalised by
midlate June. Through standardisation of databases and re-use of code,
the analyses for the CSRs, statistical reports, ISE and ISS were
produced in less than three weeks from mid June to early July. Due to
the tight timelines, project management and teamwork were key to
delivering the final submission-ready documents; the statistician and
medical writer worked closely with the client reviewers to ensure final
documents were produced and reviewed in a consistent and timely manner.
Given the delay in obtaining final and draft data, this case study
shows what can be achieved by careful planning and coordination across
teams of programmers, statisticians and medical writers without
compromising quality. |
There are many different drivers for getting the submission out of
the door as soon as possible. However, the rush can affect the review
time by authorities and may even affect the size of the label you are
requesting with the authorities. Taking a little more time, say a week,
to review all the data, information and knowledge you are supplying in
support of the label can be of great benefit. It is important to
pre-empt the questions that may come from the authorities. Most project
teams think they have supplied every eventuality in terms of summary
tables, but there is always something to be learnt from other submitted
projects.
REVIEW OF INTEGRATED SUMMARIES
The quality of the integrated summaries can have a direct impact on
the speed at which submissions are reviewed by authorities and at which
a product is introduced to the market. This can also have a direct
impact on the quality of the label approved. It is in the best
interests of the sponsor to ensure only one review cycle. It is
essential that summaries are presented in the correct way and that
information supporting the label is well organised, traceable and
understood in preparation for regulatory defence questions. The
ultimate measure for a quality submission is the number of questions
asked and the turnaround time in response. The emphasis in all cases
should be on the approval of a product and not just the submission.
If the resources are available within your organisation, put
together a plan for forming a team of experts to review the summaries
developed for your submission. They should have a critical view to try
and prompt questions that regulatory reviews potentially might have.
Pre-empting these questions could significantly improve both the review
time of a product and the reputation of your organisation in the longer
term.
Ultimately, an optimal clinical programme with great vision, design
and strategy will provide you with the expected results to support your
label, but this does need to be backed-up with the right summaries to
best explain and interpret your hard work. You may have the right data
to strengthen your target product claims, but unless the correct
information and knowledge is generated to support this, it could take
you a lot longer to get to the approval stage if your planning and
summaries are not thought through. The summaries must communicate the
vision defined through its selective programme design and expert
interpretation of the quality data generated.
Hints & tips
|
Efficacy summaries checklist
- Mention limitations of sample size
- Include age, sex, race and geographic location – clinically relevant demographic factors
- Consider US versus non-US. Does this have an affect on efficacy? Describe regional differences
- Deal with the drop-outs – planned versus actual
- Consider and discuss risk benefit
- Analyse positive and negative findings
- Focus on pre-specified endpoints
- Consider sub-populations
- Use graphical representations, such as Forest Plots
- Use consistent data formats (for example, convert to the same unit of measure)
- Use tables to combine and present data. All cells should have
something or it may be construed as missing; use consistent footnote
symbol order for every table
- When pooling data, discuss and present selection process
- State and discuss problems, which provides a more credible analysis
- Include clinical information relevant to dose recommendations and individual dose responses
- Remember that listings are not required by the FDA anymore; SAS viewer is used
Safety summaries checklist
- Choose a single dictionary, and include the dictionary and
version in the methods. If older dictionaries were used and re-coding
is not possible, include details and/or a footnote to explain
- Consistent terminology (for example, if presenting more than
five per cent common adverse events (AEs), use this cut-off throughout)
- Reference quantitative safety analysis plans (QSAPs) where applicable
- Discuss statistical issues to do with AEs; search the database for related AEs
- Always show gender (or subgroup) specific denominators
- Indicate denominator over time
- Make use of graphical presentations
- Present clinically significant criteria for laboratory, ECG,
vital signs and AEs, where applicable, referencing the most current
criteria
- For laboratory data, apply conversions where necessary to
ensure the same unit of measure for each parameter if multiple
laboratories exist within/between studies
- Ensure availability of clear documentation relating to
individual laboratory reference ranges. Lack of clarity prompts
questions around this
|
CONCLUSION
Effective planning of your integrated summaries should start early
in the process of the approval of a product. At Phase II, you should
start to plan and be prepared to adapt the plan as knowledge of the
product increases. Communicate effectively with the relevant
individuals in your teams and ensure close partnerships throughout
between your programmers and statisticians. Using metadata, define the
standards of how the data is captured, how information is derived and
stored, and how knowledge is acquired and subsequently used upfront.
This will pay dividends in the traceability of the crown jewels of your
product in the long run. Don’t rush the end game and be ready for any
questions that may come your way.
