Key Highlights

  • LB2501 produced a 100% ORR and 83.3% CR in six patients at dose level 2, but zero responses in the six patients at dose level 1, and the overall Phase 1 response rate was 50%.
  • Carvykti remains the whole business: approximately $597 million of net trade sales in the first quarter of 2026 and roughly half the economics accruing to Legend.
  • The June rally has fully reversed following a $226 million ADS offering at $29.35 and a 24 July UBS downgrade to Neutral with a $28 target.
  • Second-quarter results, expected in August, showing whether Carvykti's sequential growth from $555 million in the fourth quarter of 2025 to $597 million in the first held up.

Legend Biotech (NASDAQ:LEGN)'s experimental in vivo CAR-T therapy LB2501 produced responses in every patient treated at the higher of two doses in a 12-patient Phase 1 study, and the shares jumped roughly a third on the news in June. Seven weeks later the stock trades below where it started, after a $226 million equity raise and a broker downgrade centred on competition to Carvykti, the BCMA cell therapy that supplies almost all of the company's revenue.

On the morning of 2 June 2026, Legend Biotech released the topline of a late-breaking abstract due at the European Hematology Association congress in Stockholm. Six of six patients treated at the second dose level of LB2501, an experimental cell therapy for relapsed or refractory B-cell non-Hodgkin lymphoma, had responded. Five had complete responses. Nobody had developed neurotoxicity, and nobody had needed the chemotherapy conditioning that normally precedes a CAR-T infusion.

The stock closed that session up about 34% at $34.19, adding roughly $1.6 billion of market value on a dataset of six people. By late July it was back around $22, below where it traded the day before the abstract landed. Understanding why is a useful exercise in how much weight a very small number can bear.

What Happened

Legend presented the full LB2501 dataset on 14 June as abstract LB5006 at EHA 2026. The Phase 1 study, registered as NCT07002112, is an open-label, multicentre dose escalation. Twelve patients had been dosed as of the 1 April 2026 cutoff, six at each of two dose levels, with a median of three prior lines of therapy; 58.3% were refractory to their most recent treatment.

Across the whole cohort the objective response rate was 50% (6 of 12) and the complete response rate 41.7% (5 of 12). Every one of those responses came from the second dose level, where the rates were 100% and 83.3%. At the first dose level, none of the six patients responded. Responses spanned diffuse large B-cell, mantle cell and follicular lymphoma, and all were ongoing at cutoff. Legend's release did not state which response criteria were used or who adjudicated the scans.

On safety, the company reported no dose-limiting toxicities, no serious adverse events, no cases of immune effector cell-associated neurotoxicity syndrome and no deaths. Cytokine release syndrome occurred in 66.7% of patients and infusion-related reactions in 75%, all Grade 1 or 2, with four patients given tocilizumab and none needing corticosteroids.

Understanding the Science

Conventional CAR-T therapy is a manufacturing exercise. A patient's T cells are collected, shipped to a plant, engineered to recognise a tumour antigen, grown, tested and shipped back over several weeks. Before the cells go in, the patient receives lymphodepleting chemotherapy to clear space in the immune system.

LB2501 skips the factory. It is a lentiviral vector given as a single intravenous infusion, designed to insert the CAR gene into T cells inside the body. If that works reliably it removes the apheresis step, the manufacturing slot, the freight and the conditioning chemotherapy, turning a bespoke procedure into something closer to a drug. It targets two antigens, CD19 and CD20, rather than the single CD19 target of approved lymphoma CAR-Ts, on the rationale that tumours losing one may still be caught by the other.

Legend also reported translational data addressing the main worry with in vivo gene delivery, which is whether the vector goes where it should. Viral copy number peaked immediately after infusion and was undetectable within 24 hours; integration sites were polyclonal and diverse; and no transduction was seen in natural killer cells or other non-T lymphocytes. CAR-T cells were detectable in blood for up to 116 days.

None of this establishes long-term safety. Insertional mutagenesis risk from lentiviral vectors is measured in years, and 12 patients cannot characterise a rare event.

The Clinical and Regulatory Evidence

The honest description of LB2501 is a promising signal, not evidence of efficacy. Six patients at one dose is a cohort in which a single non-responder would have produced an 83% response rate and two would have produced 67%. The confidence interval around 100% in six patients is wide enough to be nearly uninformative about what a Phase 2 would show.

The zero-for-six result at the lower dose cuts both ways. It is consistent with a genuine dose-response relationship, which is what dose escalation is designed to find, and it is a reminder that the top-dose figure is a selected number from a study still enrolling. Primary completion is listed for mid-2027.

Legend's other early asset reported at ASCO 2026. LB2102, a DLL3-targeted CAR-T armoured with a dominant-negative TGF-beta receptor, was tested in a 20-patient first-in-human study (NCT05680922) in relapsed small cell lung cancer and large cell neuroendocrine carcinoma. Overall response rate was 20%, rising to 28.6% at the higher dose levels, with a 70% disease control rate, a median duration of response of 6.5 months, no dose-limiting toxicities or treatment-related deaths, and CRS in 30% of patients, all Grade 2 or below. Solid-tumour CAR-T has defeated many developers, and a 28.6% response rate in 14 patients is early activity rather than a franchise. Two further programmes are in or near the clinic: LB2503 in myeloma, targeting GPRC5D, and LB2505, a BCMA candidate for autoimmune disease.

The Commercial Opportunity

Everything Legend earns today comes from Carvykti (ciltacabtagene autoleucel), partnered with Johnson & Johnson's Janssen Biotech. The FDA approved it on 28 February 2022 and, after a supplement cleared on 5 April 2024, its label now covers relapsed or refractory multiple myeloma after at least one prior line, in patients refractory to lenalidomide.

That earlier-line move is what changed the economics. Carvykti net trade sales were approximately $597 million in the first quarter of 2026, up 62% year on year, with US sales up 36% and sales outside the US up 222%, according to the company's 12 May results release. Full-year 2025 sales were roughly $1.9 billion. Legend recognised $298.4 million of collaboration revenue in the quarter, close to half the reported product sales, against $175.4 million of cost of collaboration revenue.

Manufacturing was the binding constraint for most of Carvykti's life, and the metrics have improved: a 99% success rate, more than 95% on-time order releases, and availability at more than 300 sites in 18 markets. The company does not publish a slot count, but its Raritan, New Jersey plant has been described as sized for up to roughly 10,000 patients a year.

Financial Position

Legend held $834.6 million in cash, equivalents and time deposits at 31 March 2026, down from about $949 million at the end of 2025. It earned a $55 million Janssen milestone in April and raised gross proceeds of roughly $226 million on 17 June by selling 7.7 million American depositary shares at $29.35. The 30 June balance has not yet been reported.

The first quarter produced an operating loss of $49.8 million and a net loss of $54.3 million, narrower than the $101.0 million a year earlier, with an adjusted net loss of $10.5 million. Full-year 2025 net loss was $296.8 million against an accumulated deficit of $2.0 billion. Management expects to reach company-wide profitability during 2026, with cash resources extending beyond that point.

The Stock-Market Reaction

The round trip is the story. From roughly $25.50 before the LB2501 topline, the stock reached $34.19 on 2 June, then fell about 13.7% to $28.93 on 17 June when the equity offering was announced, priced below that level, and drifted lower through July. On 24 July UBS downgraded Legend to Neutral from Buy and cut its price target to $28 from $49, citing intensifying competition in myeloma from Johnson & Johnson's own bispecific antibodies Tecvayli and Talvey, the expected launch of Gilead's anitocabtagene autoleucel, and slowing sequential Carvykti growth. Shares fell about 5.6% to $22.22, leaving a market capitalisation near $4.3 billion against a 52-week range of $16.24 to $44.78. The average 12-month target among 15 analysts tracked by stockanalysis.com is $55.63, an estimate that plainly has not been refreshed against July's price action.

Key Growth Catalysts

  • Second-quarter results, expected in August, showing whether Carvykti's sequential growth from $555 million in the fourth quarter of 2025 to $597 million in the first held up.
  • A recommended Phase 2 dose and expanded cohorts for LB2501, with primary completion guided to mid-2027.
  • First clinical data from LB2503 in myeloma and LB2505 in autoimmune disease.
  • Further Janssen milestone payments, which have been lumpy but material.
  • Continued CARTITUDE readouts, including frontline studies, that could push Carvykti earlier still.

Major Risks

Carvykti's safety profile is not benign. The US prescribing information carries boxed warnings for cytokine release syndrome, ICANS, parkinsonism and Guillain-Barré syndrome, HLH/MAS, prolonged cytopenias and secondary haematological malignancies. Across 285 patients in CARTITUDE-1 and CARTITUDE-4, parkinsonism occurred in 3% and myeloid neoplasms in 5%, at a median onset of 447 days, and the label states that T-cell malignancies have occurred after treatment. On 10 October 2025 the FDA approved labelling changes adding a boxed warning for immune effector cell-associated enterocolitis, which Legend reported at ASCO 2026 at a 1.2% incidence.

Competition is the second risk, and the one the market is pricing now. Gilead acquired Arcellx in February 2026 and its BCMA candidate is expected within roughly 18 months, while bispecific antibodies are off-the-shelf and need no manufacturing slot. Third is revenue concentration: one product, one partner, roughly half the economics. Fourth is dilution, since June's raise followed a data-driven spike and more are plausible if profitability slips.

Conclusion

The strongest verified fact about Legend is not the 100% response rate. It is that Carvykti generated approximately $597 million in a quarter, that CARTITUDE-4 showed an overall survival hazard ratio of 0.55 (95% CI 0.39-0.79, p=0.001) at 33.6 months of median follow-up according to the prescribing information, and that the company is near breakeven. That is a real asset facing real competition.

LB2501 is an option on top of that, and options priced on six patients are worth something but not much. The market handing back the entire June gain looks less like a verdict on the science than a repricing of the base business. Whether the overreaction came in June, in July, or both is the question an investor has to answer.

This article is for informational purposes only and does not constitute investment advice or a recommendation to buy, sell, or hold any security. Share prices can rise or fall, and past performance does not guarantee future results. Investors should conduct independent research and consult a licensed financial adviser before making investment decisions.