

The FDA is getting a live data feed from two oncology trials run by AstraZeneca and Amgen, ditching the decades-old model of waiting months between data checkpoints. If it works, drug development timelines could shrink by years.
Imagine watching a football game, but you only get the score at halftime and at the final whistle. No play-by-play. No replays. No idea what's happening between those two moments.
That's basically how the FDA has watched clinical trials for decades. Sponsors collect data at trial sites, clean it up over weeks or months, lock it down, then ship a tidy package to regulators at a handful of predetermined checkpoints. By the time the FDA sees a safety signal or a promising efficacy trend, the moment has long passed.
Now the agency wants the live broadcast.
The FDA just kicked off its real-time clinical trials (RTCT) initiative with two proof-of-concept oncology studies: one from AstraZeneca and one from Amgen. Instead of waiting for batch data dumps at fixed intervals, the FDA will receive predefined safety and efficacy signals in near real time while these trials are actively running.
This isn't a small tweak. It's a fundamental rethinking of how regulators interact with live clinical data. And the agency's own Chief AI Officer has estimated this approach could eventually cut overall trial timelines by 20 to 40 percent. On a typical 10-year drug development journey, that's two to four years shaved off.
The technical backbone comes from Paradigm Health, whose platform transforms raw trial data into structured signals and pushes them to the FDA via API within hours of data capture. Think of it as a clinical trial dashboard that both the sponsor and the regulator can see at the same time. The FDA has already received and validated signals from AstraZeneca's trial, confirming the plumbing actually works.
The first pilot is AstraZeneca's TRAVERSE trial, a Phase 2 study in treatment-naïve mantle cell lymphoma (MCL), a type of blood cancer that remains incurable for most patients. The trial is running at MD Anderson Cancer Center and the University of Pennsylvania, testing a chemo-free cocktail of three drugs: acalabrutinib, venetoclax, and rituximab.

AbbVie is paying $10.9 billion in cash for Apogee Therapeutics, a company that was founded just four years ago. The prize: an eczema drug that could challenge Dupixent with dramatically less frequent dosing. Wall Street says the price is steep but justified.


Join thousands of biotech professionals who start their day with our free, daily briefing.
Early data from the precursor phase 1b AVR study already looks striking. Among the first 12 patients who completed AVR induction, every single one achieved a deep remission with no detectable residual disease. Across a broader group of patients still on treatment, 95.4% hit that same milestone. Six-month progression-free survival came in at 97.1%.
The trial's design makes it a natural fit for real-time monitoring. Patients who achieve deep remission get randomized to either continue treatment or stop and just be watched. If the watchers relapse, they restart therapy. That kind of response-adapted design, where the next step depends on how the patient is doing right now, benefits enormously from continuous data visibility.
The second pilot is Amgen's STREAM-SCLC, a Phase 1b study testing tarlatamab (brand name Imdelltra) in limited-stage small cell lung cancer. This is a particularly aggressive cancer where the five-year survival rate hovers around 15 to 25%, even with the best available treatment. Tarlatamab is a bispecific antibody that grabs both tumor cells and immune cells, forcing them into close combat. The trial is evaluating it after patients complete standard chemoradiation, with 12-month overall survival as the primary endpoint. Site selection is still ongoing, so no efficacy data exists yet.
Let's clear up a common misconception: the FDA is not getting a firehose of raw patient data. Nobody is livestreaming your blood work to a government server.
Instead, the FDA and each sponsor agree in advance on a reporting schema: a defined set of signals covering safety events, efficacy milestones, data quality metrics, and operational indicators like enrollment pace. These signals are aggregated, transformed, and validated before they're transmitted. Every signal can be traced back to its source data and the logic that generated it.
The difference from traditional oversight is timing and frequency. In the old model, the FDA mostly sees data through periodic safety reports, annual updates, and the occasional urgent communication. With RTCT, signals flow within days (sometimes hours) of being captured at a trial site. The agency gains what amounts to concurrent oversight rather than retrospective review.
Critically, this pilot doesn't replace existing review processes. The FDA still expects standard submissions. It still holds formal meetings with sponsors. RTCT adds a new layer of visibility on top of the traditional framework; it doesn't swap it out.
Not everyone is popping champagne. Experts have raised several legitimate concerns.
The biggest worry is dirty data. In traditional trials, sponsors spend weeks cleaning datasets before anyone at the FDA sees them. A miscoded lab value or a duplicated adverse event gets caught and corrected. With real-time reporting, those same errors could reach regulators before anyone notices the mistake, potentially triggering a false safety alarm.
There's also the question of boundary blurring. If the FDA is watching trial data unfold in real time, does it become a kind of shadow safety monitoring committee? That could complicate accountability and change the dynamic between regulators and sponsors in ways nobody has fully mapped out.
And then there's the equity problem. Building the infrastructure for real-time data pipelines, AI analytics, and API-based signal transmission requires serious resources. AstraZeneca and Amgen can afford this. A 50-person biotech in Cambridge? Maybe not. The Milken Institute flagged this directly in its public comment, warning that RTCT could widen the gap between large pharma and smaller players unless shared standards emerge.
It's no accident that both pilot trials are in cancer. Oncology is the ideal testing ground for several reasons: trials are often adaptive by design, endpoints can emerge quickly (especially in aggressive cancers), and the unmet need is enormous.
Mantle cell lymphoma has no universally curative standard of care. Limited-stage small cell lung cancer hasn't seen a meaningful systemic therapy innovation beyond platinum-etoposide in decades. Both diseases desperately need faster, smarter development pathways.
The FDA is also using this pilot to test how AI and machine learning can improve early-phase trial operations, from safety monitoring to dose selection to patient recruitment. The agency accepted public comments on its broader RTCT pilot through late June 2026 and plans to release final selection criteria in July 2026, with pilot selections completed by August 2026.
The RTCT initiative is still a pilot, not a policy change. But if it works, the implications ripple across the entire drug development ecosystem. Faster dose decisions. Earlier go/no-go calls. Shorter gaps between trial phases. And for patients with aggressive cancers, potentially years less waiting for therapies that might save their lives.
AstraZeneca and Amgen aren't just testing drugs here. They're building the infrastructure for a new way of doing clinical research, and betting that being first through the door gives them an advantage that's hard to replicate. For the rest of the industry, the question isn't whether real-time trials are coming. It's whether you'll be ready when they arrive.
SK Biopharmaceuticals just signed one of the largest AI drug discovery deals ever with Insilico Medicine, worth up to $2.5 billion. The catch? Only $18 million is guaranteed upfront, and the rest hinges on cracking one of medicine's hardest puzzles: neuroimmune disease.