Descartes-08 efficacy and safety
3 evidence topics · 11 sources
Summary: phase 2b efficacy and safety
In the phase 2b, randomized, double-blind, placebo-controlled portion of MG-001, 36 participants were randomized to Descartes-08 (n = 20) or placebo (n = 16), comprising the safety population. The primary efficacy population was a modified intention-to-treat population of 26 participants enrolled at an academic medical center with at least one postinfusion follow-up (Descartes-08 n = 15; placebo n = 11). The primary endpoint, a decrease of 5 points or more in the MGC score at month 3, was met in 10 of 15 participants (66.7%) receiving Descartes-08 and 3 of 11 (27.3%) receiving placebo (difference in proportions 0.39, 95% CI 0.01 to 0.77; P = 0.0472). In the anti-AChR antibody-positive subpopulation, the proportions were 63.6% (7/11) and 12.5% (1/8) (P = 0.0258). The company's 2024 topline release reported 71% (10/14) and 25% (3/12) in its pre-specified mITT efficacy population (p=0.018).
At month 3, mean changes from baseline in the mITT population were −5.0 versus −2.3 for MGC, −4.1 versus −1.6 for MG-ADL, −3.9 versus −1.5 for QMG, and −5.7 versus −3.7 for MG-QoL-15r. These secondary comparisons did not meet the hierarchical threshold for formal testing, and the authors attribute the lack of statistical significance in secondary outcomes to placebo responses during early follow-up. All 11 placebo participants received open-label Descartes-08 after month 3. In the Descartes-08 group, the mean (s.d.) MG-ADL reduction was −5.5 (1.1) at month 4 and −4.8 (1.4) at month 12, and the QMG reduction was −6.0 (2.1) at month 12. The share of participants with a sustained clinically meaningful response at month 12 is reported as 83.0% in the abstract and 80.0% in the Discussion; the company's April 2025 release reported 83% (10/12) of evaluable participants, defined as an MG-ADL reduction of at least 2 points. Minimum symptom expression (MG-ADL score of 1 or less) was reached by 33.3% (4/12) of Descartes-08 participants who reached month 6 and was maintained through month 12 (33.0% in the abstract). The median daily prednisone-equivalent dose in the Descartes-08 group was 9 mg at a median follow-up of 12 months, compared with 20 mg at baseline, and no participant randomized to Descartes-08 required escalation of MG therapy.
In the safety population, 17 of 20 participants (85.0%) receiving Descartes-08 and 13 of 16 (81.3%) receiving placebo reported at least one adverse event during the study, serious adverse events occurred in 5 (25.0%) and 3 (18.8%), and grade 3 adverse events in 5 (25.0%) and 2 (12.5%) participants; no grade 4 or 5 events occurred in either group for the duration of the study. The only treatment-related serious adverse event was a grade 3 infusion-related reaction with a fever of 40 °C after the fifth infusion, which resolved within 48 h without tocilizumab or steroids. Infusion-related reactions occurred in 16 (80.0%) and 9 (56.3%) participants, fever in 11 (55.0%) and 1 (6.3%), and chills in 12 (60.0%) and 0. Herpes simplex virus reactivation occurred in 3 (15.0%) and 0 participants, all with a prior history of labial herpes and on concomitant azathioprine. No case of hypogammaglobulinemia was reported. The company's April 2025 release reported no Descartes-08-related adverse events from month 4 through month 12 and no cases of cytokine release syndrome or immune effector cell-associated neurotoxicity syndrome.
Summary: phase 1b/2a and long-term follow-up
The phase 1b/2a portion of MG-001 was a prospective, open-label, multicenter, nonrandomized study in which 14 participants were enrolled (3 in Part 1 and 11 in Part 2); its primary endpoints were tolerability in Part 1 and safety in Part 2. There was no dose-limiting toxicity, cytokine release syndrome, or neurotoxicity, and 52.5 × 10⁶ CAR+ cells/kg was the maximum tolerated dose. One serious adverse event in Part 1 (grade 2 influenza after apheresis and before treatment) was unrelated to Descartes-08. Two serious adverse events occurred in Part 2: grade 3 urticaria after the third infusion, deemed possibly related and resolved after intravenous steroids, and a non-ST segment elevation myocardial infarction 72 hours after the sixth infusion in an 83-year-old man, deemed unrelated. Headache, fever, and nausea were the most common treatment-related adverse events in Part 2, in 5 (45%), 4 (36%), and 4 (36%) participants. Across 14 participants and 88 infusions, no anaphylaxis, cytokine release syndrome, immune effector cell-associated neurotoxicity syndrome, or hematologic toxicity was reported.
Among the 9 Part 2 participants in the twice-weekly and weekly arms with 12-week follow-up, mean (95% CI) changes at week 12 were −5·9 (−9 to −2·8) for MG-ADL, −7 (−11 to −3) for QMG, −14 (−19 to −9) for MGC, and −9 (−15 to −3) for MG-QoL-15r. All 7 participants in the weekly-dosing arm completed 12 months of follow-up. At month 9, their mean MG-ADL change was −6.3 (95% CI −3.5 to −9.1) and their mean MGC change −16.6. At month 12, 5 of the 7 participants maintained clinically meaningful improvement, with a mean MG-ADL change of −4.6 (95% CI −8.8 to −0.4) and a mean MGC change of −11.3 (−18.0 to −4.6). Among the 5 weekly-arm participants receiving prednisone at baseline, the median daily dose decreased from 17.5 mg to 5 mg by month 12, and neither of the 2 participants receiving IVIg at baseline required it during follow-up. No new product-related adverse events were reported during the 12-month follow-up.
Three weekly-arm participants were retreated after worsening (two at month 12 and one at month 18); two reached minimum symptom expression by week 8 and maintained it through 12 months after retreatment. In a company-reported pooled case series of 5 participants from the phase 2a and phase 2b portions (2026), the mean MG-ADL reduction from baseline at month 12 after retreatment was 6.6 (±4.2) points and the mean MGC reduction 15 (±6.2) points; no serious adverse events occurred during retreatment, and infusion-related reactions occurred with 17 of 30 retreatment infusions and 16 of 30 initial infusions.
Summary: AURORA phase 3 status
AURORA is a randomized, double-blind, placebo-controlled, multicenter phase 3 trial comparing Descartes-08 with placebo (1:1 randomization), administered as six once-weekly outpatient infusions without preconditioning chemotherapy, in adults with acetylcholine receptor autoantibody-positive generalized myasthenia gravis. The primary endpoint is the proportion of participants with an MG-ADL improvement of 3 points or more at month 4 compared with placebo; secondary endpoints include an MGC reduction of 4 points or more, QMG, MG-QoL-15r, and safety. The company states planned enrollment of approximately 100 participants; the registry record (last update posted 2026-09-01) lists actual enrollment of 128, a status of active and not recruiting, and estimated primary completion on 2026-12-30. The FDA reached written agreement on the overall trial design under the Special Protocol Assessment process in January 2025. The company expects a data readout in the first quarter of 2027 and plans a biologics license application in mid-2027; it has no products approved for sale.