U.S. flag An official website of the United States government

On Oct. 1, 2024, the FDA began implementing a reorganization impacting many parts of the agency. We are in the process of updating FDA.gov content to reflect these changes.

U.S. Food and Drug Administration
  1. Home
  2. Drugs
  3. Science and Research | Drugs
  4. Regulatory Science in Action
  5. Streamlining analysis of ion channel in vitro assays data to support clinical cardiac safety decision-making
  1. Regulatory Science in Action

Streamlining analysis of ion channel in vitro assays data to support clinical cardiac safety decision-making

Summary

CDER researchers are developing an approach that enables sharing and analyzing large volumes of cardiac electrophysiology in vitro data and streamlines the analysis and interpretation of these nonclinical data to support clinical cardiac safety decision making.

Background

The evaluation of risk to cause the abnormal and potentially fatal heart rhythm known as Torsade de Pointes (torsade) has been required for almost all new drugs since 2005. Torsade is associated with delayed cardiac repolarization, which manifests as a prolonged QT interval in the electrocardiogram (ECG). The most common mechanism for drug-induced QT prolongation is block of the ion channel encoded by the human ether-a-go-go gene (hERG) in the heart cells (Figure 1). The International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use E14 guidance (ICH E14) describes a dedicated clinical study to assess the QT prolongation potential and proarrhythmic risk of drugs, the so-called thorough QT (TQT) study.

Since 2015, ICH E14 questions and answers document (ICH E14 Q&As) R3 provides recommendations regarding use of drug concentration and QT changes from non-dedicated QT studies (e.g., first-in-human studies) as a potential substitute for a dedicated TQT study.  More recently, ICH S7B and ICH E14 Q&As released in 2022 describe how in vitro hERG studies following best practice recommendations can be used in an integrated nonclinical risk assessment to support interpretation of clinical studies assessing QT prolongation potential of drugs. This integrated approach can save time and development costs as it enables a pathway to potentially leverage existing clinical and nonclinical data without the need to conduct a separate clinical QT study to satisfy ICH E14.

Similarly to annotated ECG waveform data from clinical QT studies, it was anticipated that FDA would receive raw data of ion channel experiments as supporting information of integrated nonclinical risk assessments when implementing the new Q&As. These data would allow FDA reviewers to independently and quantitatively evaluate best practice aspects of the submitted assays regardless of whether the assays were conducted before or after the release of the Q&As. However, no open data standard existed for sharing this type of data when the most recent ICH S7B/E14 Q&As were drafted.  In addition, these data could help to understand the variability in cardiac electrophysiology in vitro study results, which is key to interpret and translate the nonclinical findings to support clinical decision making. To address these data format and knowledge gaps, CDER researchers are working together with external stakeholders.

This shows as a prolonged QT interval in the electrocardiogram (ECG), which is associated with the potentially fatal and abnormal heart rhythm Torsade de Pointes.

Figure 1. Drugs that block the hERG channel in the heart cells delay repolarization of ventricular cells. This shows as a prolonged QT interval in the electrocardiogram (ECG), which is associated with the potentially fatal and abnormal heart rhythm Torsade de Pointes. Adapted from Grilo, LS et al. 2010.

Addressing data sharing challenges through collaboration

The Comprehensive in vitro Proarrhythmia Assay (CiPA) initiative emerged in 2013 and became a global effort among regulators (US Food and Drug Administration (FDA), European Medicines Agency, Health Canada, Japan Pharmaceuticals and Medical Devices Agency), industry, and academia coordinated by multiple public–private partnerships (Health and Environmental Science Institute [HESI], the Safety Pharmacology Society [SPS], and the Cardiac Safety Research Consortium [CSRC]). The CiPA initiative proposed a new mechanistic, model-informed approach to cardiac safety assessment of new drugs, an approach made  possible by a more comprehensive understanding of the ionic currents that play a role in QT prolongation and the development of torsade.

Sharing raw data generated for CiPA research by participants around the world was challenging and time consuming because it required manual transformation of proprietary data formats into different files for subsequent analysis using additional software tools. The need for an open data format to facilitate exchange of results of ion channel experiments (i.e., cardiac electrophysiology in vitro data) for CiPA was identified and a sub-team of members from industry, academia, and CDER researchers at FDA was established to develop a CiPA Open Data (COD) format that would facilitate sharing cardiac electrophysiology data generated with either manual or high throughput automatic patch clamp systems (Figure 3, Ghasemian et. al. 2019). In addition, and to facilitate exporting existing manual patch clamp data to COD format, CDER researchers also developed the Tabulated Experimental Data (TED) format, a human friendly spreadsheet-based data format including a subset of COD elements (Mashaee et al. 2021).

When developing the COD and TED formats, the team also considered the recommendations for minimum information of cardiac electrophysiology experiments (MICEE)(Quinn et al. 2011), file types supported by electronic Common Technical Document (eCTD), and annotated waveform data formats (e.g., HL7 annotated ECG (aECG) used by the FDA ECG warehouse). As result, some of the main data elements captured in COD and TED formats include:

  • Device and software
  • Plate and well identifiers
  • Liquids, drugs, and doses
  • Protocols used (e.g., voltage, current, temperature)
  • Waveforms and measurements
    • Voltage, current, temperature
    • Raw waveform data (i.e., time series values embedded or pointing to external files)
    • Cursor locations and associated measurements

Testing the data format and analysis pipeline in study with ten experiments

Together with external stakeholders, CDER researchers also investigated analysis methods and reports frequently used to summarize results from ion channel pharmacology experiments (Gashemian et al 2019). CDER researchers also developed an automatic analysis and reporting pipeline for COD and TED datasets (Figure 2; Mashaee et al. 2021; Mashaee et al. 2024). The main analysis steps identified included:

  • linking waveforms (also referred to as traces) with test articles and concentrations;
  • flagging waveforms used for primary analysis;
  • calculating average waveform for residual current subtraction;
  • defining and evaluating measurement points (cursors);
  • calculating changes from control; and
  • modeling of dose-inhibition relationship to estimate drug potency.

     
In vitro data automatic analysis and reporting pipeline. Adapted from Mashaee et al. FDA SCDT symposium 2024.

Figure 2. In vitro data automatic analysis and reporting pipeline. Adapted from Mashaee et al. FDA SCDT symposium 2024.

The automatic analysis tool implements the following functionality of the pipeline:

  • reads from and writes raw data and results to TED or COD files;
  • optionally performs non-hERG mediated current subtraction;
  • evaluates predefined or custom cursors in each waveform; and
  • performs dose-inhibition analysis.

The reporting components of the pipeline generate tables (e.g., xlsx files) with numerical data results that can be subsequently used in other analyses or tools and interactive HTML reports with tables and figures for visual inspection of the results. These tables and figures generated by the reporting component of the pipeline include:

  • intended voltage command and cursors definitions;
  • raw waveforms;
  • current amplitude vs. time (IT) plots by cell and cursor;
  • current inhibition tables by cell and by concentration; and
  • dose-inhibition model results.

The results in the report can be used by scientist and reviewers to verify ICH S7B Q&A 2.1 best practice elements (e.g., recording quality) or to assess reproducibility of results from other analyses or summaries included in nonclinical reports of the same data. 

A first use case of the TED format and the analysis and reporting pipeline was published as supplementary materials in Alvarez-Baron et al. 2022. This study assessed hERG block potencies for 5 drugs at room temperature as well as at near physiological temperature. To promote data transparency and enable independent data verification, the complete dataset, including the original electrophysiology files (instrument readout) and plots illustrating individual cell recording quality as described in ICH S7B Q&A 2.1 best practices, were made publicly available together with TED files and source code of the analysis and reporting pipeline at https://osf.io/6w5vn/ and https://github.com/FDA/cipated (Alvarez-Baron et al. 2022). Figure 3 shows example plots generated by the pipeline from TED and/or COD format files and that would be visible to reviewers.

Schematic of a manual patch clamp system for hERG assays, showing current recordings, amplitude measurements, and dose-inhibition relationships.

Figure 3. A) Schematic of a manual patch clamp system used in hERG assays. In this schematic, solutions of vehicle-control solution and the investigational drug at different concentrations are prepared ahead and stored in reservoirs. Then, a cell expressing hERG channels is placed in a bath solution on top of a glass. Next, the cell is stimulated following a voltage protocol every few seconds (as defined per protocol) and the elicited current through the cell’s membrane is recorded with an electrode, starting with the vehicle-control solution, followed by the drug at one or multiple concentrations in an ascending fashion, and finally with a known hERG blocker (e.g., E-4031) to elicit non-hERG mediated currents. B) hERG current waveforms recorded in control (black), dofetilide 0.005 uM (orange), and non-hERG mediated current after E-4031 1 uM addition (blue). C) current amplitude measured in each recorded waveform (trace, gray area on the right of panel B) during control (black), dofetilide 0.005 uM (orange), and E-4031 1 uM (blue). D) Dose inhibition relationship is characterized using the average of the last 5 measures of each concentration evaluated in multiple cells. Adapted from Alvarez-Baron et al. 2022.

Scaling up to hundreds of experiments by nine labs around the world: HESI BAA case example

To address existing knowledge gaps in the variability of hERG as well as other cardiac ion channel in vitro results, an international effort funded by a Broad Agency Announcement (BAA) from the FDA was launched and awarded to HESI in 2019 (HESI BAA). In vitro cardiac multi ion channel data by five manual and four automated patch clamp laboratories on 28-30 drugs is being generated (Figure 4). While primary results will be based on results from analysis performed by each participant, the use of TED and COD formats for manual and automatic patch clamp data, respectively, was established early on to minimize the data sharing and analysis challenges previously faced by the CiPA initiative.

Number of drugs assessed per ion channel current, lab, and platform in the HESI BAA study (Mashaee et al. FDA SCDT symposium 2024).

Figure 4. Number of drugs assessed per ion channel current, lab, and platform in the HESI BAA study (Mashaee et al. FDA SCDT symposium 2024).

The HESI BAA effort is almost complete, and the in vitro experiments are being conducted in several phases. All participants share analysis results and intermediate associated data in tabular fashion to facilitate subsequent analyses. For example, current inhibition per cell and concentration was shared to be analyzed using the same mathematical methods and enable meta-analysis (Alvarez-Baron et al. SPS 2024, Mashaee et al. SPS 2024). In addition, manual patch clamp labs send the raw data to FDA in a rolling basis using TED format. Similarly, it is anticipated that the automated patch clamp laboratories will share their data using the most recent version of the COD format. CDER researchers used COD and TED data validation tools together with the analysis and reporting pipeline to run an independent sensitivity analysis and generate interactive HTML reports that include individual cell-level data as per ICH S7B Q&A 2.1. This approach significantly reduced the time required for performing analyses from several days to few hours, thereby increasing efficiency and ensuring reproducible results of analysis. Moreover, this allowed HESI BAA scientists to spend more time evaluating, interpreting, and discussing results. Finally, the use of these formats and tools has made it possible  to pool and analyze data from more than two hundred of hERG in vitro assay results generated under the HESI BAA providing a better understanding of the variability in cardiac electrophysiology HERG in vitro study results generated by manual (Alvarez-Baron et al. SPS 2024, Figure 5) and automated patch clamp labs (Mashaee et al. SPS 2024). 

Estimated variability of hERG assay across 5 labs and 28 drugs. See Alvarez-Baron et. al 2024 for details.

Figure 5. Estimated variability of hERG assay across 5 labs and 28 drugs. See Alvarez-Baron et. al 2024 for details.

Conclusion

CDER researchers are developing an approach that enables sharing and analyzing large volumes of cardiac electrophysiology in vitro data and streamlines the analysis and interpretation of these nonclinical data to support clinical cardiac safety decision making. Particularly, open data formats like COD and TED used in combination with data validation and automatic analysis and reporting tools developed by CDER researchers streamline the analysis of cardiac safety ion channel in vitro assay results. Moreover, these formats and tools have been used to analyze hundreds of in vitro assays conducted by HESI BAA participants around the world. Results of this analysis provide a better understanding of the variability in cardiac electrophysiology in vitro study results that are key when using these data to support clinical cardiac safety decision-making following ICH E14/S7B Q&As.

How can this work advance drug development and evaluation?

Open data formats and automated analysis tools foster collaboration, enable scientific advances based on large datasets, and streamline the independent analysis and review process of cardiac ion channel in vitro data. For example, open data formats and analysis tools developed by CDER researchers allow reviewers to perform data-driven assessment of ICH S7B Q&A 2.1 best practice elements in a high quality and time efficient fashion when hERG in vitro assay data are submitted as part of an integrated nonclinical risk assessment to support interpretation of clinical studies assessing the potential of drugs to cause dangerous cardiac arrythmias. 

Acknowledgments

This project was supported in part by appointments to the Research Fellowship Program at the OCHEN/OND/CDER and DARS/OCP/OTS/CDER U.S. Food and Drug Administration, administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. Department of Energy and FDA.

References

Alvarez-Baron et al. hERG block potencies for 5 positive control drugs obtained per ICH E14/S7B Q&As best practices: Impact of recording temperature and drug loss. J Pharmacol Toxicol Methods. 2022.

Alvarez-Baron et al. Variability of manual patch clamp hERG data generated using standardized protocols and following ICH S7B Q&A 2.1 best practices. Safety Pharmacology Society Annual Meeting. 2024 (https://hesiglobal.org/wp-content/uploads/2024/09/SPS-2024-hERG-variability-HESI-BAA_no-FDA-logo.pdf)

Ghasemian et. al. Open Format for Ion Channel Datasets from Cardiac Electrophysiology In Vitro Assays under CiPA. FDA Scientific Computing Days 2019; https://www.fda.gov/media/130504/download

Grilo et al. (2010). Stereoselective Inhibition of the hERG1 Potassium Channel. Frontiers in pharmacology, 1, 137; https://doi.org/10.3389/fphar.2010.00137 

Mashaee et al. Analysis and Reporting Pipeline for Cardiac Ion Channel Pharmacology Data. FDA Science Forum 2021; https://www.fda.gov/media/148365/download

Mashaee et al. Analysis and Reporting Pipeline for Cardiac Ion Channel Pharmacology Data: HESI BAA case example. FDA Scientific Computing and Digital Transformation Symposium. 2024.

Mashaee et al. Assessing Variability Of hERG Data Generated Using a Mock Action Potential Waveform and Automated Patch Clamp Platforms – A HESI-Coordinated, Multi-Laboratory Comparison of 28 Drugs Across 3 Platforms. Safety

Pharmacology Society Annual Meeting. 2024.
Quinn et al. Prog Biophys Mol Biol 107, 4-10 (2011).

Back to Top