In Issue 008, you watched a drug that did not exist yet move through a pipeline that took 40 years to build. We ended with daraxonrasib, a 2026 therapy targeting a cancer mutation that researchers once called undruggable. That story only exists because of something that happened between 1999 and 2001.

This is that story.

Teaser for the Root Room at the end of this Plain Talk issue. See both the Root Room track and The Informed online in our Web Only series.

The Phase III IRIS trial needed to enroll 1,106 patients with chronic myeloid leukemia (CML) across 16 countries. The planned timeline was 18 months. They enrolled in 7. That was the fastest large Phase III enrollment in oncology history at the time.

To understand why that matters, you need to understand what you are asking someone to do when you ask them to enroll into a clinical trial.

Every patient in the IRIS trial had CML. They were sick, they had options besides participating. The then-standard treatment, interferon alfa combined with cytarabine, was not ideal. It caused severe side effects, and it worked for some patients but not others. But it was something. Enrolling in a trial meant giving up the certainty of that something for a random assignment. You might get Gleevec, or you might get the interferon regimen you were hoping to avoid.

1,106 people said yes anyway.

Here is how the trial filled.

First: the drug was visibly working before Phase III even started.

Phase I Gleevec data had circulated among oncologists and patient communities. Patients with CML whose blood counts had resisted every other treatment were seeing dramatic responses. When the evidence that a drug is working spreads through a patient community, enrollment moves. People want in.

Second: patient advocacy groups already knew where their patients were.

The Leukemia and Lymphoma Society and CML-specific patient communities had infrastructure. They knew who was diagnosed, who met the eligibility criteria, and how to reach them. The trial sponsors worked directly with those networks. This is not a minor operational detail. It is one of the structural reasons the IRIS trial filled in 18 months instead of 4 years.

Third: patients found each other.

In 1999, before any social platform had a name, even Myspace, patients with CML were posting to mailing lists and early web forums. They were sharing information about the IRIS trial. They were asking each other whether anyone knew about this drug and answering from their own experience. Some of them enrolled specifically because someone they had never met told them it was worth doing.

The technology changed. The reason it works did not.

For future reference, see our free resource: Clinical Trial Registries

A drug that makes it to Phase II has already cleared Phase I, which established that it is not immediately toxic and identified a working dose range. Phase II asks a more pointed question: does this drug actually do something to the disease?

Phase III asks the harder question: does it do something better than what already exists?

This is the distinction that matters for the IRIS trial. Gleevec was not compared to a placebo. It was compared to the then-standard treatment: interferon alfa plus low-dose cytarabine. There was a reason for that, and it was not procedural. It was ethical.

When an effective treatment exists for a condition, patients in the control arm of a trial receive that treatment. Giving them a sugar pill instead would mean withholding care from sick people in order to generate data. The Declaration of Helsinki, the founding document of clinical research ethics, prohibits this when an alternative standard of care is available.

The IRIS control arm patients received real treatment. Gleevec had to prove itself against that real treatment, not against nothing.

The IRIS trial had a Data Safety Monitoring Board: an independent committee of clinicians and statisticians who reviewed the trial data at pre-specified intervals. They were not affiliated with Novartis. They did not have a financial stake in the outcome. Their job was to assess whether the trial should continue.

After an interim analysis, the board recommended stopping the IRIS trial.

Not because something went wrong.

Because Gleevec was winning so clearly that continuing would have meant keeping patients on interferon when the evidence already supported crossing over to Gleevec.

Patients in the interferon arm were offered the option to cross over to Gleevec. Most of them did.

Gleevec went from Phase I to FDA approval in approximately 3 years. COVID mRNA vaccines went from viral sequence identification to emergency use authorization in approximately 11 months.

Both had full Phase III data at the time of approval.

The question is not whether the timelines were different. They were. The question is how a shorter calendar produces the same quality of evidence. The answer is two mechanisms that had nothing to do with skipping steps.

The phases overlapped. Pfizer and Moderna ran Phase I and Phase II simultaneously, sharing safety monitoring data in real time. That is permitted under FDA guidance. It compresses calendar time without reducing patient observation or safety thresholds.

Manufacturing started before the results were in. Operation Warp Speed committed federal funding to produce doses while Phase III trials were still enrolling. If the trials had failed, those doses would have been discarded. The financial risk was accepted in exchange for time. The regulatory evidence threshold, the safety monitoring, the Phase III enrollment targets: none of that changed.

Gleevec took 3 years because the work ran in sequence. COVID vaccines took 11 months because much of it ran in parallel and the manufacturing bet was made early. Different timeline. Same steps.

Significantly, COVID vaccines did not just have the extra money pouring into them to ensure that was not an obstacle. What I found working in the industry at that time was that research sites running the vaccine trials were hoarding staff as well. They were making money enrolling vaccine trial participants, and they could pay more for clinical research coordinators and other site staff than other clinical trial sites, so naturally the vaccine trials were also populated better by staff to support the sprint speed they were running at. Similarly, many ongoing studies at the time across the entire industry were paused as the industry collectively shifted to a vaccine solution to the pandemic, which meant Contract Research Organizations (CROs) and pharmaceutical sponsors all shifted staff to those studies. Hearing from colleagues working on those studies, it was fast. Data was pouring in, and still going through the same checkpoints and safety analyses.

Two questions do the most work on the IRIS trial data.

Question 8: How many people were studied, and has this been replicated?

1,106 patients across 16 countries. Long-term follow-up data published across subsequent years, tracking patients out to 5 years and beyond. Post-marketing surveillance generated real-world outcome data from tens of thousands of CML patients treated after approval. The initial trial results were replicated in independent datasets on multiple continents. This is not a single study with a surprising finding. It is a finding that held up every time someone checked.

Question 9: What do independent experts with no stake in the outcome say?

The Data Safety Monitoring Board that recommended stopping the trial for efficacy had no financial relationship with Novartis. The FDA statistical reviewers who analyzed the application conducted their own independent analysis of the submitted data. The advisory committee that reviewed Gleevec's application voted unanimously for approval. Oncologists who had spent careers treating CML described seeing response rates unlike anything in their experience. The independent record was consistent.

Reference "The Skeptic's Toolkit" roottorx.com/field-guide

What question did this issue answer for you? What are you still wondering?

Reply to this email or leave a comment on the newsletter. Every question that comes in shapes what we cover next.

Vera Santiago is joining a clinical trial. Not the IRIS trial, rather her own.

She posted about it in the Facebook Root To Rx Lab. She was honest about why she said yes, and it had nothing to do with the drug she might receive.

Debby read the post. She had a response ready. She did not send it.

Find out what she did instead, and what a Clinical Research Coordinator Nurse (CRC) named Nani does on the first day of a trial that almost nobody knows how to run.

Read The Root Room and The Informed on the web.

New here? Start with these 3 issues.

Issue 003b: Vaccine Myths, Set Straight. The Skeptic's Toolkit — the 10 questions you will use on everything.

Issue 001: Why Your Distrust Is Rational. The founding argument. Why skepticism is healthy and how to channel it.

Issue 004: From Molecule to Medicine. The map of how drugs get built, tested, and approved.

1. O'Brien SG, Guilhot F, Larson RA, et al. Imatinib compared with interferon and low-dose cytarabine for newly diagnosed chronic-phase chronic myeloid leukemia. N Engl J Med. 2003;348(11):994-1004.

2. Druker BJ, Guilhot F, O'Brien SG, et al. Five-year follow-up of patients receiving imatinib for chronic myeloid leukemia. N Engl J Med. 2006;355(23):2408-2417.

3. Hochhaus A, Larson RA, Guilhot F, et al. Long-term outcomes of imatinib treatment for chronic myeloid leukemia. N Engl J Med. 2017;376(10):917-927.

4. World Medical Association. Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects. JAMA. 2013;310(20):2191-2194.

5. US Food and Drug Administration. Gleevec (imatinib mesylate) accelerated approval. NDA 021335. Approved May 10, 2001.

6. Polack FP, Thomas SJ, Kitchin N, et al. Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N Engl J Med. 2020;383:2603-2615.

7. Baden LR, El Sahly HM, Essink B, et al. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N Engl J Med. 2021;384:403-416.

Welcome to the Skeptical Revolution. This is a growing community learning to tell hype from evidence together, one issue at a time, and you’re part of it now. If Root to Rx isn’t in your inbox yet, subscribe at the top of this page so the next issue finds you. New here? Take the two-minute Skeptic Type quiz at openlabelmedia.com/skeptic-type to see how you read a claim. Then come think out loud with us: the Root to Rx Lab on Facebook, and @roottorx on Instagram. #SkepticalRevolution

The views expressed on Root to Rx are my own and do not represent the views or positions of my employer, or any affiliated organization.

Root to Rx | 228 Park Ave S, #29976, New York, New York 10003, United States

You are receiving this because you subscribed at roottorx.com. To unsubscribe, click the link below.