Biomarkers, pre-screened patient pools, and AI to increase productivity: Iqvia report

By Melissa Fassbender

- Last updated on GMT

(Image: Getty/elenabs)
(Image: Getty/elenabs)

Related tags IQVIA Biomarker AI Predictive analytics Clinical trials

Biomarkers have the potential to improve drug development productivity by 34% over the next five years, during which period pre-screened patient pools and the application of predictive analytics also will help address the increasing complexity of clinical trials, according to a recent report.

In 2018 the likelihood that a drug entering clinical development would be a success was 11%, down from 14% in 2017, according to a recent report published by the Iqvia Institute for Human Data Science.

Other key takeaways

According to the report, emerging biopharma companies in 2018 accounted for 72% of all late-stage pipeline activity, compared to 61% a decade ago.

Also last year, more than 1,300 life science venture capital deals were closed. These had an aggregate value of more than $23bn, which was up from approximately $10bn in deal value five years earlier, according to the National Venture Capital Association.

Large pharma companies with more than $10bn in annual pharmaceutical sales saw R&D shares drop over the last decade from 31% to 20%, per the report.

“The composite success rate for clinical development fluctuates year to year,”​ Murray Aitken, Iqvia senior vice president and executive director of the Iqvia Institute for Human Data Science, told us.

“Looking back over the past 10 years we note that the success rate moves up and down, with a long-term average that is relatively stable,”​ he said, adding that a single year decline should not be read into too much.

That being said, however, “the decline can reflect the mix of molecules, the novelty of the mechanisms being tested, and the nature of the clinical endpoints,” ​Aitken explained.

Measuring if a molecule moved to the next phase of development, success rates also can reflect a sponsor’s decision to advance a product depending on the competitive environment, the outlook for reaching the market, and expected potential economic returns, he said.

According to the report, the 2018 cohort of medicines took a median of 13.7 years from the time of first patent filing to product launch – with four new molecular entities launching in less than eight years, and 12 taking more than 20 years after the first patent filing.

Clinical trial complexity and productivity

Development timelines and productivity are a reflection of clinical trial complexity, which the report measured as a composite of subject eligibility criteria, clinical endpoints, trial sites, countries, and the number of patients participating in a trial. 

Taking all these areas into consideration, Aitken said complexity has increased from 2010 to 2018, of note for Phase I trials, as well as for oncology and immune system related trials. “We are seeing an increased number of clinical endpoints and eligibility criteria for these trials,”​ he added.

Read: Rapid increase of protocol complexity contributing to clinical trial delays

However, over the next five years, biomarker development, pre-screened patient pools, regulatory changes, as well as the increasing and expanding application of artificial intelligence (AI) and predictive analytics will affect clinical trial productivity.

“We are seeing a number of exciting approaches to clinical development that are being applied in an effort to improve success rates and R&D productivity,”​ Aitken said.

According to the report, the availability of pre-screened patient pools could reduce recruitment effort and increase success, improving productivity by 29% by 2023.

Other approaches to improve productivity include digital health technologies enabling the remote capture of drug efficacy and safety data, which Aitken said will improve patient safety, while enabling virtual trial formats and easing site burden. 

“We also see real-world data being used to optimize trial design, speed trials by aiding in investigator and site selection, and enable new trial designs by acting as virtual control arms and supporting pragmatic, adaptive and real-world evidence registry trial designs,”​ he added.

“Finally, predictive analytics and artificial intelligence ​are being used to mine data to identify new clinical hypotheses to test, reduce trial design risks and speed enrollment by identifying protocol-ready patients or predicting which patients may be eligible for a trial.”

Per the report, the application of predictive analytics could improve productivity by 16% by 2023.

The effect of these approaches, however, vary widely by therapeutic area, Aitken explained. For example, the use of biomarkers in oncology is predicted to contribute to a 71% increase in productivity, compared to 45% in cardiovascular, 25% in respiratory, and only 15% in infectious diseases.

Aitken said, “This reflects the different extents to which biomarkers are available, are able to be tested and the results analyzed accurately, and help or hinder patient recruitment.”

Overall, the report predicts the increased availability of biomarkers and related testing could improve productivity by 34% over the next five years.

Related news

Show more

Related products

show more

Saama accelerates data review processes

Saama accelerates data review processes

Content provided by Saama | 25-Mar-2024 | Infographic

In this new infographic, learn how Saama accelerates data review processes. Only Saama has AI/ML models trained for life sciences on over 300 million data...

More Data, More Insights, More Progress

More Data, More Insights, More Progress

Content provided by Saama | 04-Mar-2024 | Case Study

The sponsor’s clinical development team needed a flexible solution to quickly visualize patient and site data in a single location

Using Define-XML to build more efficient studies

Using Define-XML to build more efficient studies

Content provided by Formedix | 14-Nov-2023 | White Paper

It is commonly thought that Define-XML is simply a dataset descriptor: a way to document what datasets look like, including the names and labels of datasets...

Related suppliers

Follow us