VIR-5500: The Stealth Immunotherapy

99% Drop in Cancer Biomarkers. Really?

A post went viral on X last week claiming that a new immunotherapy drug called VIR-5500 produced "up to a 99% drop in prostate cancer biomarkers." The language was breathless. The numbers were staggering. The implication was clear: we've cracked cancer immunotherapy for solid tumors.

I need you to take a breath with me, because this story is more complicated — and more interesting — than the headline suggests.

VIR-5500 is a PSMA-targeting T-cell engager developed by Vir Biotechnology. In February 2026, the company released updated Phase 1 data from a 58-patient dose-escalation trial in men with metastatic castration-resistant prostate cancer (mCRPC) — one of the most aggressive and treatment-resistant forms of the disease.¹ Within 48 hours of the data release, Astellas Pharma announced a $1.7 billion collaboration deal to develop the drug globally.⁴

The data is real. It's also Phase 1. And those two facts don't contradict each other — they contextualize each other. So let's look at what was actually measured, how it was measured, and what it means for a disease that kills roughly 35,000 American men every year.

Teaching Your Immune System to Find the Tumor

To understand why VIR-5500 matters, you need to understand the problem it's trying to solve — and the elegant way it goes about it.

Prostate cancer has historically been invisible to the immune system. Unlike melanoma or certain lung cancers, prostate tumors create a microenvironment that suppresses immune activity. Checkpoint inhibitors — the blockbuster immunotherapy drugs that have transformed treatment for other cancers — have shown limited benefit in most prostate cancer patients.¹³ The tumor essentially hides in plain sight.

T-cell engagers (TCEs) take a different approach. Instead of releasing the brakes on the immune system (which is what checkpoint inhibitors do), TCEs physically grab a T-cell with one arm and a cancer cell with the other, forcing a direct introduction. Think of it as a molecular matchmaker that puts an assassin in the same room as the target.

VIR-5500 is what's called a bispecific antibody — one arm binds PSMA (prostate-specific membrane antigen, a protein overexpressed on the surface of prostate cancer cells) and the other binds CD3 (a receptor on T-cells). When both arms connect, the T-cell activates and kills the cancer cell.⁹

But here's the problem that has plagued every PSMA-targeting TCE before VIR-5500: PSMA isn't only on cancer cells. It's also expressed at lower levels in the salivary glands, kidneys, small intestine, and brain. Previous TCEs that targeted PSMA caused significant off-target toxicity — attacking healthy tissue alongside the tumor. Amgen's AMG 340, an earlier PSMA-targeting TCE, was discontinued after failing to show sufficient clinical activity despite manageable side effects.⁷

VIR-5500's innovation is its "dual-masked" design. Built on Vir Biotechnology's proprietary PRO-XTEN platform, the drug has both of its binding arms capped with large, inert polypeptide shields called XTEN masks. These shields physically block the drug from binding to anything — T-cells or tumor cells — while it circulates through healthy tissue.⁹

The masks are attached via linkers that are specifically cleaved by proteases (enzymes) that are abundant inside tumors but scarce in normal tissue — matrix metalloproteinases and serine proteases. When VIR-5500 reaches the tumor microenvironment, these proteases cut the linkers, the masks fall away, and the drug activates. In healthy tissue, the drug stays inert.⁹

It's a conditional activation system. The drug is essentially asleep until it reaches the right neighborhood. And the early data suggests it's working.

58 Patients, One Phase 1, and Numbers That Got Pharma's Attention

The data that broke the internet came from NCT05997615 — an open-label, non-randomized Phase 1 dose-escalation study in patients with mCRPC.³ These are men whose cancer has progressed despite hormonal therapies and, in most cases, multiple lines of prior treatment. The median patient had received four prior therapy lines. This is a population with limited options.

Fifty-eight patients were enrolled across multiple dose cohorts, ranging from 120 µg/kg to over 3,000 µg/kg administered every three weeks. The data that caught everyone's attention came from the highest-dose cohorts (≥3,000 µg/kg), which included 17 evaluable patients.^1,2

Let me unpack those numbers, because the way they've been reported matters enormously.

The headline claim — "up to 99% drop in biomarkers" — refers to the fact that 29% of patients in the highest-dose cohort (5 out of 17) saw their PSA levels fall by 99% or more. That's remarkable. But it's 5 patients. And "up to" is doing heavy lifting — it describes the best responders in the most favorable cohort at the highest dose.¹

The broader picture: 82% of patients at the highest dose saw at least a 50% PSA decline (the standard threshold for a meaningful response in prostate cancer trials). Across all dose levels, universal PSA reduction was observed — even low doses moved the needle.²

Perhaps most intriguing: 45% of RECIST-evaluable patients showed objective tumor shrinkage on imaging — not just biomarker movement, but actual measurable tumor reduction. One patient's liver metastases resolved completely.¹

The safety data is arguably the most significant finding. Across all 58 patients at all dose levels, there were zero dose-limiting toxicities and zero cases of Grade 3 or higher cytokine release syndrome (CRS).¹ CRS — an overactivation of the immune system that can cause dangerous inflammation — is the Achilles' heel of T-cell engager therapy. About 50% of patients experienced some CRS, but it was overwhelmingly mild: Grade 1 (mostly just fever) in 17% and Grade 2 in 11%.¹

For context, tebentafusp — an FDA-approved TCE for melanoma — causes Grade 2 or higher CRS in 77% of patients.¹¹ Tarlatamab, approved for small cell lung cancer, causes CRS in 55%.¹¹ VIR-5500's safety profile is a genuine standout, and it's almost certainly due to the dual-masking mechanism keeping the drug quiet outside the tumor.

A Graveyard of PSMA-Targeting Failures — and One $1.7 Billion Bet

VIR-5500 isn't entering a vacuum. PSMA has been one of the most-pursued targets in prostate cancer for decades, and the road is littered with disappointments.

Amgen's AMG 340 — an unmasked PSMA-targeting TCE — entered Phase 1 with high expectations. CRS was manageable, but efficacy was insufficient. Amgen discontinued the program and took a $650 million impairment charge.⁷ The lesson: low toxicity doesn't guarantee clinical activity. Reaching the tumor isn't enough if the drug doesn't activate potently enough once it gets there.

Amgen pivoted to xaluritamig (AMG 509), a TCE targeting STEAP1 instead of PSMA. That drug has shown a 49% PSA50 response rate overall, climbing to 59% at higher doses — respectable, but below VIR-5500's 82%.⁸ The competitive dynamics are clear: VIR-5500's Phase 1 data, if it holds, outperforms every other TCE in prostate cancer.

The most established PSMA-targeting therapy currently approved is Pluvicto (177Lu-PSMA-617) — a radioligand therapy (not an immunotherapy) that delivers targeted radiation to PSMA-expressing cells. Pluvicto has Phase 3 data showing improved overall survival in mCRPC patients, and recent PSMAddition data extended its use to earlier-stage metastatic hormone-sensitive prostate cancer.¹⁰ It's the standard against which VIR-5500 will eventually be measured — and Pluvicto has a multi-year head start in clinical evidence.

So why did Astellas bet $1.7 billion on VIR-5500 within days of Phase 1 data? Partly because the numbers are genuinely impressive. Partly because the safety profile suggests the masking technology works. And partly because immunotherapy — if it works in solid tumors — represents a fundamentally different treatment paradigm: one that could generate durable immune memory rather than requiring repeated cycles of chemotherapy or radiation.⁴

What PSA Doesn't Tell You, and What Phase 1 Can't Prove

Here's where I need to be the person at the party who asks the uncomfortable questions. Because the gap between "82% PSA response in Phase 1" and "this drug helps patients live longer" is enormous — and it's been crossed by far fewer drugs than you'd think.

PSA is a surrogate marker, not a clinical endpoint. A PSA decline tells you the drug is doing something biologically meaningful — it's reaching the tumor and affecting cancer cell activity. But PSA response doesn't automatically translate to longer survival, better quality of life, or durable disease control. Multiple drugs have shown impressive PSA responses in early trials and then failed to show benefit in Phase 3.¹³

Phase 1 trials are designed to find the right dose, not to prove the drug works. The study enrolled 58 patients with no control group, no randomization, and no blinding. Every patient knew they were receiving an experimental therapy. The "82% response rate" comes from 17 patients at the highest dose. When your denominator is 17, a difference of 2 or 3 patients swings the percentage dramatically.^1,2

The patient population was heavily pre-treated. These patients had received a median of four prior therapy lines and had limited remaining options. Response rates in earlier-line disease — where most patients are actually treated — could look very different. Vir has begun enrolling patients in combination with androgen receptor pathway inhibitors (enzalutamide and darolutamide) for earlier-line disease, but that data is not yet available.^3,5

The "99% drop" claim, while technically accurate, is statistically misleading. The viral X post stated that "29% experienced up to a 99% drop." This is true — 5 of 17 patients in the highest-dose cohort saw PSA decline by ≥99%. But framing it as a headline figure implies it's a common outcome. It's the tail end of a response distribution in the most favorable subgroup. The more representative number — 82% achieving at least a 50% decline — is still excellent for Phase 1, but it doesn't go as viral.¹

None of this means the drug doesn't work. It means the evidence stage doesn't yet support the certainty of the headlines. Phase 3 trials are planned for 2027.² Those trials — randomized, controlled, adequately powered — will answer the question that Phase 1 cannot: does VIR-5500 help patients live longer and better?

Genuinely Promising — But Not Yet Proven