Points of View Utilization of Clinical Trial Control Groups in Clinical Development

Printable PDF

Yasuhiko Nakatsuka, Senior Researcher, National Institute of Biomedical Innovation Policy

1. Introduction

The pharmaceutical industry is considering the use of "medical data" on human subjects obtained at medical institutions for various purposes. For example, in drug discovery research, it is expected to be used for target and biomarker searches, and in clinical development, for feasibility studies and patient recruitment for clinical trials, as well as for stratification of patients using Real World Data (RWD) ¹ and use as a control group for clinical trials, and for additional indications based on RWD analysis. Post-Marketing Surveillance (Post-Marketing Surveillance) In post-marketing surveillance (PMS), the use of RWD is already underway, including the collection of safety information using MID-NET.

In clinical development, too, moves toward the utilization of RWD are beginning to make progress. (For details, please refer to "How Much of the Existing Real World Data in Japan Can Be Used for Drug Development," published in April 2019 by the Clinical Evaluation Subcommittee of the Drug Evaluation Committee of the Japan Pharmaceutical Manufacturers Association (JPMA). ² ) In addition, in March 2019, the Japan Health Sciences Foundation published a report titled "Utilization of Medical Real World Data" ⁽³⁾, which discusses how to utilize medical real world data in clinical research.

The use of RWD in drug development is expected to improve the efficiency of clinical development and reduce development costs, and measures and responses are being considered in countries where new drugs are being developed, such as Europe, the U.S., and Japan. This paper focuses on the use of "clinical trial control groups" in clinical development, which is one of the themes of RWD utilization in drug development.

2. policy and guidance in the U.S.

In the United States, the 21st Century Cures ^Act4) was enacted in December 2016 to clarify the Food and Drug Administration's (FDA) regulatory scope and response to technologies such as mobile medical apps and related digital healthcare products. It also recommended the active use of Real World Evidence (RWE) ⁵⁾ as a means of easing approval regulations and mandated the FDA to support evaluation methods using RWE. Against this backdrop, the FDA recommended the active use of RWE in medical device development in August ²⁰¹⁷⁶⁾ and further issued the "Real-world evidence program " ⁷⁾ in December 2018, recommending the use of RWE to determine the efficacy and other aspects of drug products in the rare disease and cancer fields. In May 2019, the FDA issued the "Submitting Documents Using Real-World Data and Real-World Evidence to FDA for Drugs and Biologics. Submitting Documents Using Real-World Data and Real-World Evidence to FDA - Guidance for Industry " ⁸⁾ was released in draft form and is now open for public comment. The guidance is intended to help applicants for drugs that utilize RWE to provide information on the use of RWE to the FDA in a uniform format when submitting regulatory applications for approval.

The following three items are listed as examples of the information to be included regarding the use of RWE.

In the U.S., a national policy has been established to utilize RWDs and RWEs, and various guidance documents have been issued.

3. regulatory trends in Japan toward the utilization of RWD

In Japan, the utilization of RWD is also stated in the Future Investment Strategy 20189 ⁾ and the Basic Policies for Economic and Fiscal Management and Reform (Kotta-no-Moroshi 2019). ¹⁰⁾ The Clinical Innovation Network (CIN) ¹¹⁾ and the Pharmaceuticals and Medical Devices Agency's (PMDA) medical information database ( (MID-NET) ¹²⁾ to promote the utilization of RWD on a larger scale in a series of processes from drug development to safety measures. The MHLW has also issued ministerial ordinances and notifications on the utilization of RWD in drug development, including the "Revised GPSP Ministerial ^Ordinance13) " and "Basic Concept on the Use of Medical Information Databases in Post-Marketing Drug Safety ^{Surveillance14)} " The PMDA has also issued a new ordinance on the utilization of the registry. PMDA has also established a consultation framework for the concept of clinical trial planning and for improving the quality and ensuring the reliability of the registry. This consultation will be implemented on a trial basis in ^FY201915), indicating that the government is actively trying to promote the utilization of the RWD. In addition, PMDA plans to prepare a draft guideline in FY2019 that describes the basic concept of using registry data for regulatory applications and other purposes, as well as points to keep in mind regarding the reliability of data, and publish it in FY2020.

4. status of utilization of historical ^controls16) 5.

Both Japan and the U.S. have begun to develop guidelines and guidance for the use of RWD, but how many studies have actually been conducted using historical control in clinical trials? Based on the search criteria used in the aforementioned report by the Japan Pharmaceutical Manufacturers Association ( ^" To what extent can existing domestic real-world data be utilized in drug development? were applicable.

In addition, there were 20 trials in which the outcome indicator historical control was used as the control. The breakdown was as follows: Primary Outcome Measures in 9 studies, Secondary Outcome Measures in 10 studies, and Other Outcome Measures in 1 study. (Table 2)

Of these, 17 were conducted in the U.S., 1 in Denmark, and 2 were unknown. There were no trials conducted in Japan in the ClinicalTrial.gov registry.

In terms of the diseases covered by the clinical trials conducted, 18 trials were conducted in cancer, 1 trial in opioid use disorder, and 1 trial in critical limb ischemia, indicating that most of the trials were conducted in cancer. Clinical trials using historical control are being conducted for a variety of cancer types, including pharyngeal cancer, breast cancer, colon cancer, bladder cancer, non-small cell lung cancer, B-cell lymphoma, and other rare cancers.

5. movement to utilize clinical trial control groups

5-1 Flatiron Health (private company)

Flatiron Health is a vendor that maintains and operates OncoEMR, an oncology-specific cloud-based electronic medical record, providing a data platform for cancer patients and the physicians who treat them to use for treatment and research and development. Roche in 2018, but the business related to Flatiron-EHRs is independent of Roche, and services related to the EHR database can be provided outside of Roche. On this platform, the company can utilize the clinical data of a vast number of anonymized cases of approximately 2 million cancer patients.

Using the company's database, the FDA expanded the indications of Palbociclib for "male breast cancer" in April ^201917) and Entrectinib for "ROS1 fusion gene-positive locally advanced or metastatic non-small cell lung cancer (ROS1(+) NSCLC)" in August ^201918) Palbociclib (ROS1(+) NSCLC)

In addition to the Flatiron database, the IQVIA Insurance database, Pfizer's global safety database, and three RWD data sets were used as the basis for the Palbociclib (male breast cancer) application. The FDA states that "RWD tumor shrinkage and safety data" were used for the comparator group. The data used for entrectinib (ROS1(+) NSCLC) were extracted from Flatiron's database of subjects treated with crizotinib for non-small cell lung cancer and used as the basis for the regulatory submission.

How does Flatiron's database ensure data quality ^? First of all, Flatiron's database has established a procedure for acquiring clinical data to ensure "data quality" as described below.

OncoEMR, " ²⁰⁾ a component of Flatiron's platform, is a cloud-based electronic medical record specialized for cancer diseases, covering 275 medical institutions that provide cancer treatment. OncoEMR, which has been installed at each medical institution, allows originally structured data such as laboratory test values and blood pressure to be standardized within OncoEMR, and consistent data is accumulated as a database. Medical information in "OncoEMR" includes unstructured summaries and other descriptions by physicians. In order to structure such unstructured data, a person in charge called an "abstractor," who is trained to maintain data quality, is assigned to each medical institution. clinical information that needs to be structured (e.g., cancer stages obtained from text data).

By following these procedures, the company is building a medical database that can be used for approval ^{applications21)}.

5-2 Medidata case study (private company)

Medidata is currently working to match individual patient data that meet the criteria for a new clinical trial for acute myeloid leukemia from the control group data accumulated on the platform from more than 3,000 trials using Medidata's Rave EDC (Electronic Data Capture) system to date. The company is working to create a Synthesis Control Arm (SCA) by matching individual patient data that meets the criteria for a new clinical trial for acute myeloid leukemia from the more than 3,000 control arms accumulated on its platform using the Electronic Data Capture (ECDC) ^system22).

The company has verified that the outcome data of SCAs created using this platform are comparable to the data of the actual control arms in randomized controlled trials conducted in the past for non-lung small cell ^carcinoma23). Unlike existing RWDs, the database in this case consists of clinical trial data from each pharmaceutical company using Medidata's Rave, which Medidata integrates to create composite control groups for a variety of diseases. The clinical trial data will be anonymized. The use of clinical trial data is expected to ensure that the quality of the data will be high enough to be used for regulatory applications. It is also expected that if the disease is such that the Outcome to be obtained is specified and standardized, it will be possible to obtain the Outcome data required in the clinical trial.

5-3 Practical feasibility study of a control group

In February 2019, the FDA and Flatiron Health announced an expansion of a collaboration signed in 2016 to utilize cancer RWD from medical health electronic records (EHRs).

Additionally, in April 2019, the FDA and Brigham and Women's Hospital announced a demonstration project using RWE to predict the outcomes of seven ongoing Phase IV Randomized Clinical Trials (RCTs) (RCT DUPLICATE This project will use RWE (insurance databases, electronic medical records, etc.) to predict the results of Phase IV trials ⁽ RCTs) that have not yet been concluded, before they are completed, and verify the results of the RCTs against the predicted results analyzed by RWE to see if similar judgments can be made. The RWE will be used to verify whether the same decisions can be made.

Interim results are scheduled to be reported in mid-2019 and final results at the end of 2020.

Given the fact that the FDA is actively verifying whether or not RWE can be used as an alternative to RCTs in collaboration with private companies, if these verifications yield favorable results that allow RWE to replace RCTs, a trend toward RWE replacing RCTs under certain conditions is likely to occur.

5. summary

In fact, the clinical trials that have used historical data have mainly focused on oncology, and it is expected that the number of trials that utilize historical control, including those for rare diseases, will increase in the future. This is because it is difficult to use placebo as a comparator in clinical trials for cancer and rare diseases, and it is also difficult to set up a control group in clinical trials for rare diseases (including rare cancer diseases) because the number of eligible subjects is small. However, it is difficult to use an existing RWD as a control group for a clinical trial because of various issues such as the quality of the RWD (in terms of consistency, completeness, and accuracy), the possibility of obtaining outcome data, and the influence of bias when using an existing RWD as the basis for data for regulatory applications.

In Japan, it has been proposed to promote the use of CIN registries for clinical trial control groups, and in the "Report of the Study Group on Utilization of Big Data in the Medical and Health Fields, Vol. 3," published by the National Institute of Health and Welfare ^Policy25), the construction and utilization of disease registries suitable for secondary use was promoted through the development of electronic medical record templates The report recommends the construction and utilization of medical data using electronic medical record templates as a means of promoting the construction and utilization of disease registries suitable for secondary use. While considering the use of disease registries, it is necessary to consider the construction of a unified platform (electronic medical record, etc.) that can obtain highly reliable data, such as that of Flatiron, including private companies. If the construction of such a database will promote the utilization of clinical trial control groups, pharmaceutical companies will be able to reduce the cost and time spent on clinical trials, and furthermore, eliminate the number of subjects assigned to receive placebo in RCTs.

Return to News List

TOP