A first-in-human phase 1 trial has found the therapy can produce substantial, and sometimes drug-free, responses in patients with severe rheumatoid arthritis that has resisted multiple treatments.
CD19 CAR T-cell therapy has driven down disease activity and RA-associated autoantibodies in six patients with severe, treatment-refractory rheumatoid arthritis, while producing no serious adverse events.
The phase 1 results of the prospective COMPARE trial, published in Nature Medicine, provide early proof-of-concept evidence that CAR T therapy may offer a new approach for a small group of patients with difficult-to-treat RA.
“These findings extend anecdotal reports and provide systematic, trial-level data showing that this approach is both feasible and safe,” the researchers wrote.
“It additionally provides proof-of-concept evidence for clinically meaningful responses to CD19 CAR T-cell therapy in a subset of patients with difficult-to-treat, long-standing, treatment-refractory RA.”
All six participants had long-standing, active, ACPA-positive RA and had previously failed numerous therapies. They had received a median of two conventional synthetic DMARDs and 6.5 biologic or targeted synthetic DMARDs. Four had previously received rituximab.
After stopping DMARD treatment and undergoing lymphodepletion, the three women and three men received a single infusion of the fully human autologous CD19 CAR T cell therapy mivocabtagene autoleucel, or miv-cel. Follow-up ranged from 36 to 52 weeks.
Disease activity fell in every patient. DAS28-CRP declined by a median 49% to week 24 and by 34% at the latest follow-up.
Four of the six patients achieved DAS28-CRP remission at some point, with three maintaining remission without DMARDs or glucocorticoids throughout follow-up.
At week 24, five patients had achieved an ACR20 response, four an ACR50 response, and two an ACR70 response. By the latest follow-up, three had achieved an ACR70 response.
The clinical response was backed by imaging. Ultrasound showed marked reductions in synovial inflammation, including a median 100% reduction in synovial hypervascularisation at the latest assessment.
“The study achieved its primary endpoint, which demonstrated that miv-cel therapy had an acceptable short-term safety profile in a heavily pretreated population of patients with RA,” the researchers wrote.
But the researchers also found evidence that the treatment was doing more than suppressing inflammation.
CAR T therapy rapidly depleted circulating CD19-positive B cells and also eliminated CD19-positive B cells from bone marrow, synovial tissue, and lymph nodes – including tissue-resident cells that had persisted in patients previously treated with rituximab.
RA-specific autoantibodies subsequently plummeted. ACPA against mutated citrullinated vimentin fell by a median 94%, with four of six patients becoming seronegative. Anti-CCP ACPA fell by a median 91%, while rheumatoid factor IgM and IgA declined by 89% and 96%, respectively.
Five of six patients became seronegative for RF-IgM and four of six for RF-IgA.
Related
When B cells eventually returned in four patients, the reconstituted population consisted predominantly of naive and transitional B cells rather than the antigen-experienced cells seen before treatment.
ACPA-expressing memory B cells were initially absent, leading the researchers to suggest CAR T therapy had effectively eliminated the autoreactive B-cell memory compartment.
That finding supports the possibility of an “immune reset”, rather than simply another form of ongoing immunosuppression.
The response was not universal or necessarily permanent, however. Two patients had only partial clinical responses, one restarting baricitinib at week 36 because of persistent disease activity. Another patient relapsed after 44 weeks following a period of drug-free remission.
The relapsing patient also showed early B-cell reconstitution and later re-emergence of autoreactive plasmablasts, providing a possible biological clue to recurrence, the researchers said.
They stressed that such observations were hypothesis-generating given the cohort of just six patients.
Safety was the primary focus of the phase 1 study. All six patients developed cytokine release syndrome, but cases were limited to grades 1 or 2 and resolved after treatment.
There were no cases of immune effector cell-associated neurotoxicity syndrome and no serious adverse events.
One dose-limiting toxicity, a grade 3 elevation in liver transaminases, resolved without sequelae. Cytopenias were common following lymphodepletion, while infections were infrequent and predominantly mild.
“Although all patients developed low-grade CRS, these events were easy to manage, the researchers wrote.
“Importantly, no high-grade CRS and no ICANS or LICATS was observed. Cytopenia after LD was universal, but largely self-limited or manageable with G-CSF support.”
The researchers said that longer follow-up and larger studies would be needed to establish the risk-benefit profile, particularly given potential late infections, cytopenias, secondary malignancies, and prolonged B-cell depletion.
However, they said the findings provided trial-level evidence that CD19 CAR T-cell therapy was feasible in refractory RA and could produce clinically meaningful, treatment-free disease control in some patients.
“In summary, our findings provide initial safety data and proof-of-concept that CD19 CAR T-cell therapy may induce clinically meaningful responses and enable achievement of a drug-free sustained remission in a subset of patients with highly refractory RA,” the researchers concluded.
“If confirmed in larger controlled studies, CD19 CAR T-cell therapy may represent a distinct therapeutic approach for a selected group of patients with RA.”
The phase 1 safety findings have allowed the COMPARE study to advance to phase 2.
