One immune-cell infusion held myeloma at bay 5 years in half of 20 patients

Half of a small group of people with multiple myeloma, a blood cancer of the bone marrow, were still alive with no sign of their cancer growing five years after a single infusion of the cell therapy Carvykti, and none of them needed more myeloma treatment in that time.¹ The 20 patients had already been through 1 to 3 earlier rounds of treatment, earlier in their illness than the people Carvykti was first tested in.¹ The results come from a trial called CARTITUDE-2 and were presented on September 25, 2026, at the International Myeloma Society’s annual meeting in Glasgow, Scotland.¹

What happened

Carvykti, also known by its generic name cilta-cel (ciltacabtagene autoleucel), is a CAR-T therapy. In this kind of treatment, immune cells called T cells are given a new receptor so they can find and attack a target on cancer cells. Carvykti’s target is a marker on myeloma cells called BCMA (B-cell maturation antigen).¹ It is given once, as an infusion.¹

The new report covers “cohort A,” the first group of patients in CARTITUDE-2, a phase 2 study (a mid-stage trial) that tests Carvykti in several different groups of myeloma patients.² All 20 people in this group had myeloma that had come back after 1 to 3 earlier treatment regimens, and their cancer had kept growing despite lenalidomide (Revlimid), a standard myeloma pill.¹ They had also received drugs from two other common myeloma drug families before.¹ On average they had had 2 earlier regimens, and 85% had already had a stem cell transplant using their own cells.¹

Patients got no maintenance treatment after the infusion.³

The results so far

By the time of this analysis, patients had been followed for a median of 60.7 months, about five years.¹ The main numbers reported were:

• 10 of 20 (50%) were alive, with no cancer growth and no further myeloma treatment, at least five years after the infusion.¹
• 19 of 20 (95%) had a complete response at their best point, meaning standard tests could no longer find the cancer.¹ The remaining patient had only a small response and died of COVID-19 pneumonia.²
• Half of the patients went at least 60.5 months before their cancer grew or they died. The estimated share still free of cancer growth at five years was 54.2%.¹
• The estimated share still alive at five years was 69.2%.¹

The other 10 patients did less well. Seven saw their cancer grow or died of it before five years, one left the study without the cancer growing, and two died of side effects before five years.²

The researchers also looked for patterns. People still free of cancer growth at five years tended to start with less cancer in their bone marrow: a median of 20% of marrow cells were myeloma-type cells, compared with 40% among those whose cancer grew.¹ They also had higher levels of certain helper T cells, both when cells were collected and when the CAR-T cells peaked in the blood.¹ Long-lasting control was also seen in some people with genetic features that usually mean a harder-to-treat cancer: of the 7 long-term responders with genetic data, 5 had at least one such high-risk feature.¹ These were exploratory comparisons in a small group.³

What it does not show yet

This was a single group of 20 treated patients, with no comparison group given another treatment. AJMC noted the analysis is limited by its small size and single-arm design.¹ Lead researcher Adam Cohen of the University of Pennsylvania also cautioned that the group is small and that long-term follow-up from a larger trial, CARTITUDE-4, is needed to confirm the finding.⁴

The results show that the cancer did not grow, not that it was gone. The study’s plan did not require a bone marrow test at five years. Only 3 of the long-term responders had one, and all 3 showed no detectable cancer at a very sensitive level of one cancer cell in a million.¹ The trial’s main goal was this kind of deep “no detectable disease” result, measured earlier; how long people stayed free of cancer growth and how long they lived were exploratory measures, not the trial’s main questions.²

The findings were presented as a poster at a medical meeting.² The trial registry lists CARTITUDE-2 as still running, with no results posted there yet.⁵

Safety

With longer follow-up, no new nerve-related side effects from the CAR-T cells were reported, including facial nerve problems or Parkinson’s-like movement symptoms.¹ Since the previous report at about 30 months, 5 patients (25%) developed second cancers, which were judged unrelated to Carvykti.¹ These were acute myeloid leukemia, breast cancer, prostate cancer, two kinds of skin cancer and an early melanoma; one patient had two of them.¹ There were two new deaths: one from myeloma, and one from breast cancer in a patient whose myeloma had not grown in five years.¹

Who is behind it

Carvykti was developed by Johnson & Johnson and Legend Biotech.¹ CARTITUDE-2 is sponsored by Janssen Research & Development, part of Johnson & Johnson, and has enrolled 208 people across all its groups since it started in November 2019.⁵

The lead author, Cohen, and coauthors wrote:

“Only cilta-cel has demonstrated evidence of durable 5-year treatment-free remissions in [relapsed/refractory multiple myeloma], now with 50% of patients in remission in a CARTITUDE-4–like population,”

That statement was reported by OncLive.² AJMC quoted Niels van de Donk, a hematology professor at Amsterdam University Medical Center, from a statement, saying the findings add evidence that using effective therapies earlier may give more patients “the opportunity to achieve deep, durable remissions and long-term disease control without maintenance.”¹

Some of the coverage comes from education sites funded by drugmakers. The Multiple Myeloma Hub, which interviewed study coauthor María-Victoria Mateos, lists Johnson & Johnson and Legend Biotech among its sponsors, and its article on this study was supported by Legend Biotech; the site says funders have no influence on content.³

How it fits the field

The new data build on CARTITUDE-1, an earlier study in people whose myeloma had already been through at least 3 earlier treatment regimens.² In that study, 33% (32 of 97) were alive and free of cancer growth at least five years after one infusion.¹ Median time before cancer growth was 33.3 months there, compared with 60.5 months in the new group.²

Researchers read the two studies together as a hint that giving the therapy earlier may help more people reach long-lasting remission.¹ Matthew Frigault of Mass General Brigham made the same point, noting that about one-third of the heavily treated CARTITUDE-1 patients were free of cancer growth at five years.⁴ But the two studies enrolled different people at different times and were not designed to be compared directly, and the investigators described the gap as “numerically higher” results rather than a tested difference.²

Carvykti is already approved for earlier use. In April 2024 the FDA approved it for adults whose myeloma came back after at least one earlier treatment and no longer responds to lenalidomide.¹ That approval was supported by the larger CARTITUDE-4 trial, which enrolled patients similar to this group.⁴ Researchers are also testing the therapy even earlier: in a separate study called CAR-PRISM, 20 people with high-risk smoldering myeloma, an early form that has not yet caused symptoms, had no detectable disease after one infusion and none had progressed to active myeloma after a median of 15.3 months.⁶

What to watch next

• Longer follow-up. After the study group closed, patients were invited into CARTinue, a follow-up study that tracks people for 15 years after their infusion; 14 of the 20 enrolled.²
• CARTITUDE-4 long-term results. Cohen said these are needed to confirm whether the five-year pattern holds in a larger group.⁴
• Late side effects. Cohen said Parkinson’s-like symptoms may be less frequent in less heavily treated patients, and that lowering the amount of cancer with other drugs before the infusion has been linked to a lower risk.⁴
• The rest of CARTITUDE-2. The trial’s main data collection is scheduled to finish in February 2027.⁵