Persistent CAR-T Cell Infiltration and JAK-STAT Reprogramming Drive Enterocolitis After BCMA CAR-T Therapy
Seventeen patients developed severe intestinal inflammation after receiving ciltacabtagene autoleucel, the BCMA-targeted CAR-T cell therapy for multiple myeloma, and researchers have now traced the damage to the patients’ own engineered T cells colonizing the gut.
The findings, published in Nature, describe CAR-positive T cell infiltration of the intestinal mucosa in a condition the investigators call CAR-T cell–associated enterocolitis, or CAR-TEC, Nature reported.
Among 160 patients who received cilta-cel infusions between July 2022 and October 2025, 17 underwent endoscopy for possible enterocolitis, according to the study. Most cases emerged within six months of infusion. One patient was not diagnosed until nearly 700 days after treatment, a detail that extends the risk window well beyond what clinicians had assumed.
- Severe enterocolitis after BCMA CAR-T therapy appears driven by the engineered CAR-T cells themselves infiltrating the gut, not by off-target inflammation alone.
- In a cohort of 160 patients receiving cilta-cel infusions between July 2022 and October 2025, 17 underwent endoscopy for possible enterocolitis (159 infusions occurred at MD Anderson, of which 158 were consecutive between July 2022 and August 2025); several cases were fatal or refractory to multiple drugs.
When the therapy becomes the disease
Ciltacabtagene autoleucel reprograms a patient’s T cells to hunt cells carrying B cell maturation antigen, a marker on myeloma cells.
The same target sits on normal B cells, and the new study documents what that means for the gut: profound depletion of mucosal B cells and plasma cells, paired with expansion of highly cytotoxic CAR-T cells, in biopsy tissue from 10 patients with CAR-TEC, compared with 7 CAR-T–treated controls without enterocolitis and 26 healthy volunteers, per the study summary.
The analysis combined single-cell transcriptomics, flow cytometry and tissue imaging. Beyond the immune depletion, the team found inflammatory remodeling of myeloid, stromal and glial cells, with interferon signaling and Janus kinase–associated reprogramming across stromal, endothelial and epithelial compartments. That pattern, the researchers reported, supports JAK inhibition as a mechanism-based therapeutic strategy rather than an empiric guess.
How the cases unfolded
The clinical picture that emerges from the Nature paper is sobering. In certain instances, the antecedent infection involved detection of C. difficile DNA with negative endotoxin results, yet every patient nonetheless received antibiotic treatment targeting C. difficile.
Median onset of cases was day +61, with one patient having symptoms that began near day +541 and evaluated at approximately day +700 at an outside facility. Symptoms typically prompted a workup 27 days later, and patients had cycled through a median of 6 unique agents per patient (range 1–7) by the time a regimen controlled the disease.
Seven patients received steroids—usually prednisone or dexamethasone—starting a median of 31 days after symptom onset without achieving effective disease control (median of 6 days on steroids, range 2–34 days); applied locally, mesalamine and budesonide were given to 8 patients for a median of 73 days (range 13–218 days).
At our institution, remission was achieved through T cell directed therapies, namely cyclophosphamide or calcineurin inhibitors. Cyclophosphamide was administered to four patients at doses ranging from 1 to 2 g m− 2; two of them responded rapidly with no relapse of the enterocolitis, although one later experienced myeloma relapse and another died soon after from infection, with cyclophosphamide immune suppression a likely contributor to that death.
Two patients were treated with the calcineurin inhibitor cyclosporine and had prompt responses, including Patient 3, who had failed to respond to multiple agents including cyclophosphamide. Patient 147 achieved a complete clinical response on intravenous cyclosporine, and biopsies demonstrated a drop in BCMA+ CAR infiltration from 25% of CD3+ cells to 5% of CD3+ cells, as shown by clinical flow cytometry.
Two patients received ruxolitinib as a primary agent and one patient as a maintenance agent. Neither patient treated with ruxolitinib as primary therapy responded, and one died of hemorrhagic C. difficile infection.
The single patient given ruxolitinib as maintenance, following remission induced by intravenous (IV) cyclosporine, ultimately relapsed with IEC-EC while on oral ruxolitinib, as the CAR+ population expanded again to 27% of BCMA+ CAR—once more documented by clinical flow cytometry.
Mortality followed the disease’s severity. In one patient, symptoms started around day +541, and evaluation at an outside facility took place at roughly day +700. Another, treated only with steroids after a late diagnosis, presented in cardiac arrest from polymicrobial bacteremia.
A third patient succumbed to an infection on day +468 even though infliximab and ustekinumab brought symptomatic improvement; longitudinal staining of GI biopsies taken at day +217 and day +457 afterward revealed ongoing CAR+ T cell infiltration.
Two groups, one mechanism
The Nature paper and the study describe the same syndrome from complementary angles. The Nature report centers on the MD Anderson clinical cohort — 159 infusions at MD Anderson, of which 158 were consecutive between July 2022 and August 2025, 17 patients with possible IEC-EC who underwent endoscopy (16 identified prospectively and 1 retrospectively) — and the treatment histories behind them.
As laid out in the publication summary, the contribution establishes the framing of CAR-TEC and supplies the interventional evidence missing from the clinical series: upadacitinib—an oral selective JAK 1 inhibitor—produced clinical, endoscopic and histologic improvement in two people with CAR-TEC.
Where the accounts differ is emphasis. The Nature material stresses treatment heterogeneity and the poor outcomes of late diagnosis; the summary stresses the mucosal biology and positions CAR-TEC within the broader expansion of B cell–targeted therapies into autoimmune disease, where gut surveillance questions will only grow. Both agree on the core finding: the engineered cells persist in tissue and drive the damage.
The research was carried out in accordance with the Declaration of Helsinki and received approval from the Institutional Review Board of MD Anderson Cancer Center (MDA protocol nos. 2025-1294, 2022-0042 and 2005-0656). Stanford biospecimen collection, flow cytometry analysis and clinical data collection were approved under Stanford protocol no.
CCT5019, cleared by the Stanford Institutional Review Board, and participants provided written informed consent before enrollment in the study covering biological sample collection, clinical data collection and genomic data analysis, per the paper’s ethics statement.
What this changes in the clinic
For myeloma patients on cilta-cel, the practical message is that persistent or severe diarrhea weeks to months — and in rare cases years — after infusion warrants endoscopic evaluation rather than empiric infection treatment alone, given that infection workups delayed diagnosis in four cases.
The finding that T cell–directed agents and calcineurin inhibitors produced the durable remissions gives clinicians a hierarchy where none existed; the JAK-signaling data give a second, mechanism-anchored option backed by two documented upadacitinib responses.
Hematologists weighing cilta-cel can review the peer-reviewed evidence base on BCMA-directed therapy toxicity as it accumulates.
The open question the data leave: with 17 patients with possible IEC-EC identified in a single center’s cohort and at least one diagnosis approaching two years post-infusion, how many centers are biopsying early enough — and whether JAK inhibition, tested in only two patients so far, will hold up as a standard of care. The study’s own treatment table shows the answer is still being written, one case at a time.
Disclaimer: The information provided in this article is for educational and scientific communication purposes only and does not constitute medical advice. Always consult with a qualified healthcare provider regarding any medical condition, diagnosis, or treatment plan.