Drug makers that run clinical statistics in both SAS and R now have a vendor-supported way to do it in one place. SAS and Posit PBC, the company formerly called RStudio, said on Nov. 16 that they have partnered to integrate SAS Viya and SAS Clinical Acceleration with Posit Workbench, Positron, Posit Connect and Posit Package Manager, with clinical development as the first focus.
What SAS and Posit announced
The SAS announcement describes SAS Clinical Acceleration as a product built on SAS Viya that combines a clinical data repository, a statistical computing environment and AI-assisted clinical data transformations to prepare research for regulatory submission. Posit’s tools are meant to extend it so that teams can bring open source code into a managed environment.
One piece works already, according to Posit. “Today, a programmer can write SAS and R code in Positron, and SAS Viya runs the SAS code under that programmer’s own permissions,” said Eric Pite, Senior Vice President of Product and Strategy at Posit, in the announcement. He added: “We are working with our joint customers to decide which clinical workflows to support next.”
The announcement says customers can work in managed, versioned R environments alongside SAS Clinical Acceleration, that Posit Package Manager helps keep R and Python environments consistent and reproducible, and that teams can develop code with Positron and Posit Assistant. It gives no price, no product version numbers and no date for the next set of workflows.
The customer named in the release
Regeneron is the one customer quoted. “Using R and SAS to tackle parts of the same workflow used to mean workarounds, copied data and duplicate access control governance,” said Dilip Raghunathan, Business Solutions Partner at Regeneron. “Now our programmers write both in one session under their own access rights, with AI assistance that works within those same controls.”
By that account, the AI assistant runs with the access rights of the person using it. Platform teams are applying the same per-user principle to coding agents. We covered it yesterday, when AWS and Databricks detailed how to meter coding agents by user and team.
Why mixed SAS and R teams are the target
Posit publishes its own figures on how far clinical programming groups have moved toward R. Its pharma page says GSK moved half of its clinical programming from SAS to R, and that Novo Nordisk went from a pilot R project to a complete R-based new drug application accepted by the FDA in under two years. The same page says all of the top 20 pharma companies use Posit for clinical analytics. These are Posit’s own claims about its customers.
SAS had already opened its side. The SAS Clinical Acceleration product page says users can perform analyses using SAS, R or Python in a governed environment. The partnership adds Posit’s own editor, package manager and publishing server, connected to Viya.
A Posit customer story about an unnamed top 10 pharmaceutical company shows the cost of running the two worlds apart. SAS programmers who wanted to adopt R or Python had to learn a new language and a different interface at the same time, the story says. That company runs two platforms, an exploratory one and a GxP platform with validated containers for clinical reporting.
Open tools arriving with a vendor’s terms attached is a pattern in models as well. In October Mistral released Large 4 weights with a separate license for model-as-a-service.
What the FDA pilots show
The regulator has reviewed submissions written entirely in R. The R Consortium, a non-profit, reported that its Pilot 3 test submission was an all-R package for ADaM datasets and for tables, listings and figures. It went through the FDA’s eCTD gateway on Aug. 28, 2023, was resubmitted on Apr. 19, 2024 after agency feedback, and received a final FDA response letter on Aug. 8, 2024. The consortium calls the work an FDA and industry collaboration and has published the materials.
Two earlier pilots covered R scripts for tables, listings and figures, and R Shiny code. In our reading, those pilots explain the shape of this partnership. A sponsor can file in R, so SAS has reason to keep R work inside its governed platform, and a sponsor with years of validated SAS programs has reason to keep running them.
What it means for the buying committee
Heads of statistical programming, clinical data platform owners and quality leads will each have a different question for the two vendors.
Access and audit
Pite’s statement covers SAS code, which Viya runs under the programmer’s own permissions. SAS says its repository supports roles, permissions, electronic signatures, version control and audit trails, and Posit says Workbench maintains audit trails from code to submission. A mixed session touches both products. Ask which system holds the record an inspector would read, and whether R steps and AI-assisted edits appear in it.
Package qualification
The announcement says Package Manager keeps environments consistent and reproducible. It does not say who qualifies each R or Python package for regulated use. That work stays with the sponsor’s validation group unless a contract says otherwise.
Scope and cost
Only one workflow is described as available today, and the rest are still being chosen with joint customers. The announcement does not say whether the integration needs additional licenses from either company. A purchase order should name the workflow being bought and the release it ships in.
What to do this quarter
Teams that already license both products can test the claim directly. Run one study program that mixes SAS and R in Positron, then pull the access log and the audit record from each system and compare them. Teams choosing a platform should ask both vendors for the list of supported clinical workflows in writing, with dates. SAS and Posit staff are at PHUSE EU Connect in Glasgow through Nov. 18, at booths P3 and 43, according to the announcement.
Source: SAS Newsroom