Conducting a clinical study takes a considerable amount of planning, resource, and commitment. But, after the last subject is out, the database has been cleaned, and the planned analyses have been done, what happens to the data that were collected?
The data from clinical studies may be required by regulatory authorities for a variety of purposes, such as to support marketing authorisation applications or renewals, or to provide additional information on marketed drugs. For interventional studies conducted in accordance with the International Conference on Harmonisation (ICH) Guideline for Good Clinical Practice (GCP),
1 data collected in a clinical study should be submitted to the authorities in a Clinical Study Report (CSR).
This basic introduction to the CSR covers:
- What a CSR is
- What a CSR is used for
- What a typical phase III CSR contains
- Who writes the CSR
- How your input into the CSR may be needed!
What is a CSR?
The CSR is a detailed picture of a clinical study. Why was it done? What was done? What were the results? Most importantly, the CSR assesses whether the study answered the questions (objectives) it was designed to investigate.
The format and content of a CSR depends on the study design, which in turn depends on the stage of development of the study drug. Typically, phase I CSRs contain pharmacokinetic and safety data, phase II CSRs contain information on dose selection and safety, and phase III CSRs contain confirmatory efficacy and safety data. A CSR may be over 200 pages long, and that does not include the appendices, which can stretch to 1000s of pages. Fortunately, CSRs can now be submitted to the authorities electronically, which makes submission logistically more manageable, and saves on paper.
CSRs intended for submission to regulatory authorities should be compliant with ICH Guideline E3,
2 which specifies the content and layout. This guideline is used for most phase II and III studies, with the format adapted for earlier and later phase studies. A CSR adhering to ICH E3 guidelines should be acceptable to regulatory authorities of all countries and regions who have adopted ICH guidelines, making it easier for pharmaceutical companies to submit their data around the world. However, some regions do have specific requirements that must also be considered. For example, for studies to be included in a marketing application in the European Union (EU), the principal investigator must sign off the CSR to confirm that they have read it and that it accurately describes the study and results; for other regions, the sponsor’s responsible medical officer may provide the only signature.
Shortened CSRs, called abbreviated CSRs, may be acceptable to the authorities in some cases, eg, uncontrolled studies. In abbreviated CSRs, the presentation of data can be brief and some sections specified in the ICH E3 guideline may not be needed at all, but a full description of safety data is still required. For studies not intended for submission to regulatory authorities, CSRs do not need to be ICH E3 compliant, although they typically follow the same basic format.
What is a CSR for?
In many cases, the CSR is the building block of a marketing application or renewal. This application/renewal may be sent to multiple regulatory authorities around the world, depending on a pharmaceutical company’s marketing strategy. CSRs provide the raw data from which a regulatory authority may decide to license a drug (or not); anyone who is involved in a clinical study or production of the CSR can therefore impact the likelihood of a drug being marketed.
Typically, a marketing application will comprise multiple CSRs, often spanning years of research from the first phase I studies in healthy volunteers to phase III studies in large numbers of patients. The data presented in the application are directly used to justify the proposed prescribing information of a drug. The data from the individual CSRs are compiled into summary documents such as clinical summaries and clinical overviews, which look at all the studies together and give an overview of the clinical programme. The summary documents assist the authorities in assessing the application, eg, in terms of the balance of the efficacy versus safety of the drug.
The data in a CSR may also be used as the basis for manuscripts for publication in the medical literature, for the purpose of updating health professionals on developments in the field. In early phases of the clinical development programme, the results in CSRs are used to update the Investigator Brochure to ensure investigators and ethics committees have the current data on a drug’s development. For marketed products, CSR safety data will be included in pharmacovigilance reports such as Periodic Safety Update Reports, which assess the safety of products after marketing. Also, for Food and Drug Administration (FDA)-approved drugs (and soon for clinical studies conducted in the EU), key data from a clinical study must now be included in a public area such as www.ClinicalTrials.gov, and the CSR is the best place from which to extract these data into any necessary summaries.
What is included in a CSR?
The main ICH E3 CSR sections for a phase III study are summarised below:
Synopsis
The synopsis comprises a concise overview of the whole CSR. It describes key features such as the study design, the study population, the planned statistical analyses, and the key results and conclusions of the study.
Methods
The methodology section of the CSR describes the planned conduct of the study as presented in the protocol (and any protocol amendments) and the statistical analysis plan, and any changes from that plan. Importantly, this section describes the objectives of the study and how specific data variables were collected and analysed to meet these objectives. The statistical analyses of the study data are specified before the analysis is started in the statistical analysis plan; this limits the analyses that are done to prevent “data mining” for trends that might not be statistically valid.
If there were any problems in the way the data were collected this would be described and an assessment made as to how this affected the results. This section also describes the criteria for enrolling subjects in the study, details of study drug administration and which medications were prohibited during the study.
Subject disposition
This section includes an account of the number of subjects who entered and completed the study, usually by treatment group. It is important to know whether more subjects dropped out in one treatment group as this might bias the results, especially if they dropped out because the drug was not effective or because of its side effects. The reasons for subject withdrawal are also presented, again to check for bias (eg, is one reason more common in one treatment group?)
Protocol deviations
The protocol deviation section includes a summary of the major protocol deviations, usually by treatment group. It is important to know whether major deviations, such as use of a medication prohibited by the protocol, occurred with a similar frequency in each treatment group or whether there was a possible bias in one group that might affect the validity of the results. If major protocol deviations are very significant, especially if present in only one treatment group or one site/country, this could obviously affect data interpretation and even whether the study could be considered GCP-compliant.
Demography & baseline characteristics
This section includes a summary of the characteristics of the subjects in the study. Demographic characteristics are subject age, sex, ethnicity, body mass index etc. Baseline characteristics refer to previous medications taken, concomitant medications, and prior and concurrent medical history; it may also include characteristics relevant to the particular subject population that might affect responses to study drug, eg, cancer status in oncology studies. This section enables the CSR reviewer to assess whether the subjects recruited into the study represent the subjects that the drug is eventually intended for. Importantly, in randomized studies, similar characteristics seen across treatment groups indicate a lack of bias and that randomization was successful.
Efficacy
The results! For phase III studies this section is typically split into the analysis of primary and secondary efficacy endpoints. Endpoints are simply the data variables that have been collected at specific time points, eg, blood pressure at Day 30.
The data are typically analysed by treatment group, with comparisons between treatment groups where there are more than one group (ie, comparing those subjects who received study drug with those subjects who received placebo).
This section tends to contain the majority of any statistical analyses conducted, such as analyses to see if there were statistically significant differences between treatment groups. Analysis of the primary efficacy endpoint is the most important; in phase III studies it is often called the confirmatory analysis and directly relates to the primary objective of the study, eg, was the study drug better than placebo at achieving a specific endpoint?
Safety
The safety section includes a detailed discussion of adverse events observed during the study, usually by analysing the events by incidence, intensity (mild, moderate or severe), and relationship to study drug (ie, were the events considered related to study drug administration or not). Narratives of serious adverse events and other significant adverse events, such as those causing discontinuation of study drug, are presented. Narratives are a description of the relevant adverse event and its outcome put into context of the subject’s previous medical history, their experience in the study, and other adverse events they have experienced in the study.
All other safety variables collected are also usually presented in the safety section, eg, laboratory data such as haematology parameters and vital signs such as blood pressure. The main aim of this section is to assess the safety profile of the study drug, often by comparing it to placebo or a comparator: was it as expected or were there any new safety concerns; is the safety profile better than that of the comparator drug? Extent of exposure to study drug is also described in the safety section. For example, did the majority of subjects receive the amount of study drug required by the protocol? If not (ie, if they received the wrong dose or were not treated for long enough), this could affect how well the study can assess the safety (and efficacy) profile of the study drug.
Discussion & conclusions
The discussion section summarises how well the study data supported the study objectives as well as the clinical relevance of the results. Importantly, any data limitations are also described (eg, if the study failed to recruit the planned number of subjects, the main efficacy analysis may not be sufficiently robust to make claims of statistical differences between treatment groups). The main conclusions of the study are presented in this section.
End of text summary tabulations & figures
All summary tables and figures are presented in this section. If there were additional analyses that were not pre-specified, these are also presented.
Appendices
The appendices to a CSR can be very large. These comprise all documents needed to support the CSR, including the protocol, sample informed consent form, sample case report form (CRF), statistical analysis plan, administrative details, and subject data listings. The data listings typically include all relevant data collected from the CRF, and the appendices may include copies of CRFs for those subjects with serious adverse events (the data for whom the authorities may be particularly interested in).
Do I need to provide input?
You might do: many personnel involved in the running of the study or with specific expertise will need to provide input into the CSR.
Statisticians and data managers will usually be responsible for providing the statistical output and data listings to the CSR author, without which CSR writing cannot begin. They can also advise the author on the best way to interpret and present complex statistical analyses. Personnel most familiar with how the study was conducted (ie, project managers and clinical operations staff) may be asked to specify how and why the study conduct deviated in any way from the study protocol (or, hopefully, confirm that it did not!). Experts in regulatory compliance can provide feedback on whether the CSR is ICH E3 compliant and whether there are specific requirements for a regulatory authority in the country or region to which the marketing application will be submitted. Medical experts may review a CSR to ensure that the clinical relevance of results is interpreted correctly. The appendices to a CSR comprise many of the core documents of a trial master file, so clinical research associates (CRAs) may be asked to provide these documents (eg, curriculum vitae of investigators, laboratory accreditation certificates, CRFs).
Who prepares the CSR?
Writing a CSR requires knowledge of the relevant regulatory guidelines and an ability to present the study data in the most appropriate way for the intended audience. The author also needs to coordinate the input from the relevant experts and be adept at managing any conflicting opinions on the CSR! Specific technical expertise is also needed to ensure the electronic appendices to the CSR meet regulatory requirements. Increasingly, medical writers (experts in clinical regulatory writing) are called upon to prepare the CSR, although clinical operations staff, project managers or statisticians may also write CSRs in some companies. Due to the complexities of the CSR, its preparation is a lengthy business, typically taking several weeks, but the importance of this document make the production of a quality CSR critical.
Summary
After a clinical study is over, the data collected are typically presented to the regulatory authorities in a CSR. The CSR is a large and complex document, that describes whether the study met the objectives it was planned to investigate. Even once your main role in the study is over, the CSR author may need your input, whether to provide source documents or review the CSR content. Ultimately, production of a high quality CSR should be considered the final point in the clinical study, not last subject out, or database lock, and the CSR author may need your help!
References
- Guideline for Good Clinical Practice E6(R1). International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use: ICH Tripartite Guideline; 1996.
- Structure and Content of Clinical Study Reports E3. International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use: ICH Tripartite Guideline; 1995.
