Spyre SPY003: A Promising Readout That Still Isn't Randomized

Spyre Therapeutics just filed another positive data point across its inflammatory-bowel-disease pipeline, and the stock carries the record of a winner: shares sit near $89, up roughly 170% year to date, against a 52-week range that started around $14.50. The newest headline drug, SPY003, is an extended half-life anti-IL-23 antibody, and its early mid-stage results hint at the "best-in-class" positioning the company is chasing. Before the euphoria sets in, read the small print on the study—because the number that matters most is still a year away.

What the readout actually measured

SPY003 was tested in Part A of a Phase 2 study called SKYLINE, in 44 patients with moderately to severely active ulcerative colitis. The study's main measure was a tissue-level inflammation score called the Robarts Histopathology Index, and patients improved by an average of 10.0 points from a baseline of 17.2 at week 12, a result the company called statistically significant. Secondary numbers were more modest on their face: 20% of patients reached clinical remission by the modified Mayo score, 30% showed endoscopic improvement, and the average modified Mayo score fell 3.5 points.

Here is the caveat that should govern how you read all of it. Part A was an open-label, single-arm assessment—a single dose level, no placebo group, no blinding. The primary endpoint compares each patient against their own baseline, which is why it could be "statistically significant" in a 44-person study without a control arm. That is not the same thing as proving the drug beats doing nothing or beats an existing treatment.

The company knows this. Part B of SKYLINE, currently enrolling, is the randomized, placebo-controlled test of monotherapies and combinations—including the pairwise combos SPY120, SPY130, and SPY230. That is where the drug meets a real comparator, and topline induction data are not expected until 2027.

Putting 20% and 30% in perspective

The temptation is to read the 20% clinical remission and 30% endoscopic improvement as weak because they look like smaller versions of the marketed IL-23 drugs. A cleaner reading comes from checking against the placebo rate those marketed drugs were measured against. In the Phase 3 study that supported approval of risankizumab (Skyrizi), an IL-23 antibody in the same class, induction produced a 20.3% clinical remission rate versus 6.2% for placebo, and a 36.5% endoscopic improvement rate versus 12.1% for placebo.

So Spyre's absolute 20% and 30% figures, measured in an open-label study of a harder-to-treat population (41% had already failed an advanced therapy), sit clearly above the level where you'd expect placebo to land on its own. Treat the numbers as encouraging but unanchored: because there was no simultaneous placebo arm, you cannot compute a treatment effect, only the absolute rate.

That same caveat dulls the "best-in-class" claim the company attaches to SPY003. Compare within Spyre's own platform: SPY001 (anti-α4β7) produced a 9.2-point RHI reduction and SPY002 (anti-TL1A) a 10.7-point reduction from the same open-label design. The three sit tightly together, which is a good sign for the platform and an awkward one for any single-drug supremacy story.

Why this one drug is only part of the thesis

The reason the market pays attention is not one molecule. Spyre's antibodies are engineered with an extended half-life—SPY001, for example, reported a half-life above 90 days, supporting dosing every three to six months rather than the frequent infusions patients on marketed drugs face. The combination program is the deeper bet: putting two confirmatory targets into a single injection with infrequent subcutaneous dosing, positioned against a market that already has well-established, approved IL-23 drugs.

That means today's SPY003 readout functions less as proof of one drug and more as incremental evidence that the whole pipeline works. It also means the 2027 Part B platform readout—which covers the combos and the dose ranging—is the single event that can confirm or kill the larger story, not the September data point alone.

The price already reflects a lot of the promise

This is where structure and price part ways. Spyre holds about $1.15 billion in cash as of June 30 and says that funds operations into the second half of 2029, so the clinical story is well-financed. Yet the stock's roughly $7.8 billion market capitalization has already absorbed much of the "the platform works" thesis through its 170% year-to-date run. The next twelve months bring several binary catalysts of their own—the rheumatoid arthritis sub-study due this month, psoriatic arthritis and axial spondyloarthritis reads in the fourth quarter, and then the long wait to the controlled 2027 data.

The discipline here is the same as it would be anywhere a real mechanism meets a big price. The mechanism is defensible: the extended half-life and combination angle are real engineering bets, and the open-label numbers sit above typical placebo rates. But the study that can actually confirm SPY003's place against the marketed IL-23 class does not report for another year. A 44-patient, single-dose, unblinded snapshot is an encouraging checkpoint, not a verdict—and at roughly 90 times the stock's starting price for the year, the market has decided to pay for the confirmation before it arrives.

The investable question is whether you're comfortable underwriting a year of waiting and a cascade of future readouts for a result that today is genuinely promising but not yet proven against placebo. For some, the answer is the platform itself has now cleared its cheapest screen. For others, the definitive test being priced in before it exists is exactly the point to stay on the sidelines.