Finding 144 bone lesions instead of 99 changes how doctors see the disease
In the long effort to make cancer visible before it becomes untreatable, a first-in-human trial presented at the 2026 Society of Nuclear Medicine and Molecular Imaging meeting has offered a new way of seeing. Researchers at the University of California, San Francisco tested a radioactive imaging agent that seeks out a protein called CD46 on prostate cancer cells — a target that persists even when other molecular markers fade — and found it revealed far more disease than conventional scans. For men with the most resistant form of prostate cancer, the ability to see what was previously hidden is not merely a technical advance; it is a step toward knowing where the enemy truly is.
- Metastatic castration-resistant prostate cancer can outpace current imaging tools, leaving oncologists navigating with incomplete maps as the disease spreads.
- The new PET agent, 89Zr-DFO-YS5, uncovered 144 bone lesions where a standard bone scan found only 99, and 33 soft tissue sites where CT imaging caught just 14 — a detection gap that could change treatment decisions.
- Thirty patients completed the trial safely, and a premedication protocol nearly eliminated infusion reactions, clearing a key safety hurdle for broader use.
- A preliminary signal suggests patients whose tumors absorbed more of the imaging agent responded better to a paired CD46-targeted drug therapy, hinting at a future where the scan predicts the cure.
- The next trials will pit this agent directly against PSMA PET — the current gold standard — to determine whether CD46 imaging fills the gaps left when cancer evolves beyond PSMA expression.
At the 2026 Society of Nuclear Medicine and Molecular Imaging annual meeting, researchers unveiled the first human results for a novel PET imaging agent designed to find prostate cancer through a different molecular door. The agent, 89Zr-DFO-YS5, targets CD46, a protein that sits on the surface of prostate cancer cells and — crucially — remains present even in tumors that have lost PSMA, the protein that most modern imaging tests depend on.
Felicia Tang and her team at UCSF had already developed the underlying antibody, YS5, and demonstrated its safety as a drug. For this study, they attached a radioactive tracer to it and enrolled 30 men with metastatic castration-resistant prostate cancer between April 2022 and November 2025. Patients were scanned in three sequential cohorts to optimize timing, dosing, and image quality. The five-to-seven-day post-injection window emerged as the sweet spot for imaging clarity.
Safety was manageable. Radiation exposure was comparable to other zirconium-89 agents, and a premedication regimen reduced infusion reactions to just one case among the final 24 patients. An attempt to improve image quality by adding unlabeled antibody backfired — it increased reactions without improving scans — and was dropped.
The detection numbers were striking. In individual patients, the new PET found substantially more bone lesions and soft tissue metastases than conventional bone scan and CT imaging. Tumor uptake was strong across the cohort, with clear separation between cancer tissue and background signal.
Twenty-three patients subsequently received a CD46-targeted antibody drug conjugate alongside standard hormone therapy. Those who responded — defined by a significant PSA drop — had modestly higher tumor uptake on their baseline PET scan than non-responders, a trend that did not yet reach statistical significance but pointed toward a potential predictive role for the imaging agent.
The road ahead involves direct comparison with PSMA PET and formal testing of whether 89Zr-DFO-YS5 can reliably identify patients most likely to benefit from CD46-targeted treatment. If those studies hold, oncologists may gain both a complementary detection tool and a way to match patients to therapies before committing to a course of treatment.
At the 2026 Society of Nuclear Medicine and Molecular Imaging annual meeting, researchers presented the first results from a human trial of a new way to find prostate cancer tumors. The imaging agent, called 89Zr-DFO-YS5, targets a protein called CD46 that appears on the surface of prostate cancer cells. What makes this approach different is that it works regardless of whether the cancer has lost expression of PSMA, a protein that other modern imaging tests rely on. For men with metastatic castration-resistant prostate cancer—the most aggressive form—having another imaging option could matter.
Felicia Tang of the University of California, San Francisco, led the work. Her team had previously developed an antibody called YS5 that binds to CD46, and they had already tested a drug version of it in patients, showing it was safe and had some anti-cancer activity. Now they took that same antibody, attached it to a radioactive tracer, and tested whether it could light up tumors on a PET scan. The trial enrolled 30 men with metastatic castration-resistant prostate cancer between April 2022 and November 2025. Patients received an injection of the imaging agent and underwent scanning. The researchers divided them into three groups: the first three patients got multiple scans over time to measure radiation exposure; the next three got a single scan plus an extra dose of unlabeled antibody to see if that improved image quality; and the remaining 24 got a single scan at five to seven days after injection, which earlier data suggested was optimal timing.
The safety profile was acceptable. The radiation dose was comparable to other zirconium-89 based PET agents, with the highest absorption in the adrenal glands, spleen, liver, and kidneys. When researchers gave patients a premedication regimen before the injection, infusion reactions dropped dramatically—only one patient in the final cohort of 24 experienced a reaction. The extra unlabeled antibody given to the second group did not improve image quality and actually increased infusion reactions, so that approach was abandoned.
What stood out was the imaging performance. The new PET scan detected 144 bone lesions in one patient compared to 99 found on a conventional bone scan. In another patient, it found 33 soft tissue sites—lymph nodes, liver spots, lung nodules—versus only 14 on CT imaging. Across all 30 patients, the average uptake in tumors was strong, with a mean SUVmax of 11.70, a measure of how much radioactivity accumulated in cancer tissue relative to normal tissue. The tumor-to-blood ratio averaged 9.78, indicating clear separation between cancer and background.
Twenty-three of the patients went on to receive a drug called FOR46, a CD46-targeted antibody drug conjugate, combined with enzalutamide, a standard prostate cancer hormone therapy. Researchers tracked how well these patients responded by measuring PSA levels. Those who achieved a 50 percent drop in PSA had a slightly higher tumor-to-blood ratio on their baseline PET scan—9.6 compared to 7.6 in non-responders—though this difference did not reach statistical significance. The trend suggests the imaging might eventually help predict who will benefit most from CD46-targeted therapy, though more data is needed.
The implications are still being worked out. Prostate cancer imaging has been transformed in recent years by PSMA PET scans, which have become the standard for detecting metastases in men with advanced disease. But PSMA expression can be lost as cancer evolves, leaving some patients without a good imaging option. CD46 appears to be expressed more consistently across different prostate cancers, independent of PSMA status. The next phase of research will directly compare 89Zr-DFO-YS5 to PSMA PET to see where each excels, and will test whether the imaging can reliably predict which patients will respond to CD46-targeted drugs. If those trials confirm what this first-in-human study suggests, oncologists may soon have a complementary tool for finding hidden metastases and matching patients to the therapies most likely to work.
Notable Quotes
CD46-targeted PET imaging with 89Zr-DFO-YS5 was well tolerated and demonstrated high tumor uptake in men with mCRPC— Felicia Tang, University of California – San Francisco
Compared to conventional imaging, 89Zr-DFO-YS5 PET identified additional lesions, supporting its potential clinical value— Felicia Tang, University of California – San Francisco
The Hearth Conversation Another angle on the story
Why does it matter that this imaging agent targets CD46 instead of PSMA, which doctors already use?
Because PSMA expression isn't stable. As prostate cancer progresses and evolves, some tumors lose PSMA. When that happens, PSMA PET scans become less useful. CD46 appears to stick around across different cancer types and stages, so it's a more reliable target.
The study found 144 bone lesions versus 99 on bone scan. That's a 45 percent improvement. Is that clinically meaningful?
It is, because finding more lesions changes how doctors stage the disease and plan treatment. If you're missing 45 lesions, you're underestimating the extent of spread. That affects prognosis and which therapies make sense.
One patient had only one infusion reaction out of 24 in the final group. That seems very safe.
It does, especially after they added premedication. The earlier cohorts had more reactions, but once they adjusted the protocol, it became manageable. That's how phase one trials work—you learn and refine.
The PSA response trend wasn't statistically significant. Does that mean the imaging can't predict who will respond to treatment?
Not yet. The trend was there—responders had higher tumor-to-blood ratios—but with only 23 patients and a p-value of 0.053, you can't draw firm conclusions. Larger trials will tell us if this is real or noise.
What happens next?
Head-to-head comparison with PSMA PET in more patients. If 89Zr-DFO-YS5 finds lesions PSMA misses, and if it can predict treatment response, it becomes a tool oncologists actually use. Right now it's promising, but still experimental.