Investigational ADC Shows Encouraging Activity in CEA-High Advanced NSCLC

In a conversation with CancerNetwork^®, Matthew Lee, MD, an assistant professor of medical oncology at City of Hope in Duarte, California, about the first clinical disclosure of BG-C477, an investigational antibody-drug conjugate (ADC) that pairs a CEA (CEACAM5)–targeting antibody with a topoisomerase 1 inhibitor payload, presented at the 2026 European Society for Medical Oncology (ESMO) Congress.^1,2 In the monotherapy dose-escalation and safety-expansion cohorts of a phase 1a study (NCT06596473), 43 patients with heavily pretreated non–small cell lung cancer (NSCLC) were treated with BG-C477 at doses of 2.7 to 3.6 mg/kg every 3 weeks; 25 patients (58.1%) had CEA-high expression, defined as immunohistochemistry (IHC) 2+ or 3+ in at least 50% of tumor cells.¹

As of the September 10, 2026, data cutoff, the confirmed objective response rate (ORR) was 60.0% (95% CI, 38.7-78.9) in the CEA-high subgroup, with a disease control rate (DCR) of 92.0% (95% CI, 74.0%-99.0%), compared with an ORR of 37.2% (95% CI, 23.0%-53.3%) in the overall NSCLC population. Treatment-emergent adverse events led to treatment discontinuation in 4.7% of patients.¹ BG-C477 received FDA Fast Track Designation in October 2026.²

Lee discussed what differentiates BG-C477’s design from other topoisomerase 1 inhibitor–based ADCs, how confident he is that CEACAM5 expression will hold up as a biomarker for patient selection, how its toxicity profile compares with other ADCs in NSCLC, and where it might fit into the broader ADC sequencing landscape. He said the preliminary safety, tolerability, and efficacy data for BG-C477 support CEACAM5 as a potential actionable biomarker in lung cancer, though he cautioned that these dose-escalation, hypothesis-generating findings must be backed by larger expansion and later-phase data.

CancerNetwork: BG-C477 targets CEACAM5 via a topoisomerase I inhibitor payload, joining an increasingly crowded ADC space in NSCLC. What differentiates its design or early clinical profile from other topoisomerase 1 inhibitor–based ADCs already in development or approved?

Lee: One of the main aspects of this ADC, BG-C477, is that it targets CEA with a proprietary type of antibody that’s been shown in preclinical studies to be very selective, with a stable, cleavable linker and a topoisomerase 1 inhibitor payload. But it also [has] an enhanced internalization component, so it goes into the cells and targets the cancer cells themselves while also creating a bystander effect, similar to some other ADCs, but in a more selective, targeted way.

Its toxicity profile also reflects that selectivity, in the sense that it’s targeted to go against the CEACAM5 biomarker. It was designed to be more selective as an antibody, with a stable, cleavable linker, but also to still be effective.

The CEA-high subgroup showed a higher ORR vs the CEA-low group. How confident are you that CEA/CEACAM5 expression will hold up as a clinically actionable biomarker for patient selection, and what would validation of that biomarker require?

This is a first-in-human, phase 1 type of trial, so the numbers presented right now are very preliminary, but they do show a promising signal. From the early numbers we’re seeing, the CEA-high group, which is IHC 2 to 3-plus in greater than 50% of tumor cells, shows a differential response compared with the [lower-expressing group].

Initially, the phase 1 design used a serum CEA cutoff of greater than 5 ng/mL, and then we retrospectively looked at CEA IHC and saw a differential response. We’re also interested in whether serum CEA correlates with CEA IHC, and that’s one of the key things we’ll [look at] as we gather more information in the dose-expansion phase and get larger numbers to see whether this pans out. At this moment, with the preliminary data we have, I’d say the CEA IHC 2 to 3-plus group is probably the most promising group to look at for now, but it will be interesting to see whether we can also incorporate serum CEA in the future, which could further expand how many patients might benefit from this [CEACAM5-targeted] ADC.

How does the toxicity profile of BG-C477 compare with what you’ve seen from other ADCs in NSCLC, and what does it suggest about how it might be sequenced or combined?

Compared with what we’ve seen in the past, one of the main things is that this CEACAM5 ADC does show some selectivity in how it’s being used as a topoisomerase 1 inhibitor payload, and a lot of the [adverse] effects we’ve seen in this early-phase trial are really related to cytopenias and gastrointestinal [adverse] effects.

The majority of these [adverse] effects haven’t led to much dose reduction or treatment discontinuation; only a small proportion of patients discontinued treatment because of these effects.Overall, it does show that this is a tolerable CEACAM5 ADC, and that there’s promise it can move forward in the sense that it’s not only safe but also potentially efficacious. I think it also highlights that while these topoisomerase 1 inhibitor payloads are driving some of these toxicities, it’s not leading to the same rates of toxicities we’ve seen with other topoisomerase 1 inhibitor ADCs, probably because of where we’re targeting with CEA. Some other agents have shown interstitial lung disease or pneumonitis, and those reported rates are not what we’re seeing in this early-phase trial, which is very important, especially for [patients with] lung cancer and lung cancer providers.

With multiple ADC platforms now competing for the post-chemotherapy, post-targeted-therapy niche in NSCLC, how do you envision BG-C477 fitting into that sequencing landscape if later-phase data confirm this activity?

That’s the key question: where does this fit into the whole landscape of ADCs? If you think about what’s currently approved, every ADC basically has a biomarker-driven approach, whether that’s MET IHC with telisotuzumab vedotin [Emrelis], datopotamab deruxtecan [Datroway] specifically in EGFR-mutant [disease], or trastuzumab deruxtecan [Enhertu] in HER2-mutant NSCLC.

One of the major things about this trial is that it’s similarly looking at CEA by IHC, but in a broader sense, it has included patients with both EGFR-mutant and EGFR wild-type disease. Some of the trials are now looking at whether to expand into the wild-type vs mutated populations; right now, they’re leaning more toward EGFR wild-type [disease]. But I think the other future idea is to see how this could expand into other potential antigens. For now, what we’re seeing is efficacy in both groups from the early data, which could make this a broader type of ADC compared with some other options.

What’s the next step for this program in terms of dose optimization, expansion cohorts, or combination strategies that you would want to see before this moves toward a registrational trial?

Right now, they just finished enrollment for dose optimization, so we’ll have to see how those data look and where the recommended phase 2 dose can go from there. Those are some of the future plans for a single-arm phase 2 trial to see if this could potentially support accelerated approval in the EGFR wild-type lung cancer population.

There are also potential plans for a phase 3 trial, which is also in the design stage, possibly compared with a standard-of-care option like docetaxel. These trials aren’t open or accruing yet, but these are the kinds of designs they’re hoping to pursue in the future for this asset.

What would you want practicing oncologists to take away from this first disclosure of BG-C477 data, given how early and hypothesis-generating a dose-escalation cohort is?

One of the key points is that CEA has been looked at historically in lung cancer, but it hasn’t really been targeted in a successful, efficacious, and safe way until this point. This is promising in terms of the design of the drug, its safety and tolerability, and even the preliminary efficacy, which really does hold some promise that CEACAM5 can be a potential new biomarker in lung cancer.

We knew about MET and HER2 in the past, and we’ve had some familiarity with CEA, but mainly in the serum, not as a way of targeting. We’ve been using it as a monitoring and prognostic [marker]. This is the first time we can try to target through protein expression by IHC and see some actual efficacious benefit.

References

1. Lee M, Gong Y, Fu S, et al. BG-C477, a CEA (CEACAM5)-targeting antibody-drug conjugate, in patients with advanced non-small cell lung cancer: results from a phase 1a dose-escalation and safety-expansion study. Presented at: European Society for Medical Oncology (ESMO) Congress 2026; October 23-27, 2026; Madrid, Spain. Abstract 2567P.
2. BeOne presents compelling data in hard-to-treat liver and lung cancers at ESMO 2026. News release. BeOne Medicines Ltd. October 26, 2026. Accessed October 28, 2026. https://tinyurl.com/u5y6utsf