SUMMARY  

• Johnson & Johnson does not recommend the use of CARVYKTI in a manner that is inconsistent with the approved labeling.
• Several studies have evaluated CARVYKTI outcomes in the context of out of specification (OOS) vs in-specification products, across multiple prior lines of therapy (LOTs), and in relation to cellular composition and clinical factors influencing manufacturing outcomes^1-4:
  • Bowden et al (2025)² presented CARVYKTI manufacturing outcomes across multiple prior LOTs.
  • Dingli et al (2025)³ presented CARVYKTI outcomes for OOS products, comparing safety and efficacy with in-specification products.
  • Vucinic et al (2024)⁴ presented real-word data on factors influencing autologous lymphocyte collection and subsequent manufacturing of CARVYKTI or idecabtagene vicleucel (ide-cel) in patients with refractory or relapsed multiple myeloma (RRMM), focusing on cellular composition and T cell senescence.

CLINICAL DATA

CARVYKTI Manufacturing Outcomes Across Multiple Prior LOTs

Bowden et al (2025)² presented CARVYKTI manufacturing outcomes across multiple prior LOTs.

Study Design/Methods

• Commercial CARVYKTI manufacturing data from July 2024 to October 2025 were evaluated to determine factors influencing manufacturing outcomes, identify biomarkers associated with manufacturing success, and assess manufacturing results after OOS-related remanufacturing or recollection.
  • Manufacturing success is defined as the successful production of chimeric antigen receptor (CAR) T-cell products that comply with specifications.

Results

Manufacturing Outcomes: OOS Rates

• Lower OOS rates were observed when T cells were sourced from patients with fewer prior LOTs (7.8% with 1 prior LOT vs 10.8% with ≥4 prior LOTs; N=4576) as presented in Figure: OOS Rates
• Net OOS rate was 8.8% (95% confidence interval [CI], 7.8%-9.8%).

Biomarkers Predictive of Manufacturing Success

• Fewer prior LOTs were associated with favorable recovery of cluster of differentiation (CD) 3+ cells at the start of manufacturing as compared with ≥4 prior LOTs.
• The initial post-thaw CD4:CD8 ratio improved by 30% in patients with fewer prior LOTs compared with those with ≥4 LOTs.
• Viable cell recovery prior to lentiviral transduction and cumulative population doubling level (cPDL) improved by up to 6% in patients with fewer prior LOTs.
• Activation bag recovery was improved with 1 or 2 prior LOTs vs ≥4 prior LOTs.
• Markers of metabolic activity, such as glucose consumption and lactate production, significantly improved with fewer prior LOTs (average improvement of 10%) for patients with 1 prior LOT vs more prior LOTs.

CARVYKTI Manufacturing Outcomes

From All Prior LOTs (N=385)

• At the first attempt:
  • A total of 64.7% (n=249) were released through an expanded access program, without remanufacturing.
  • Among the initial OOS products, 28% of products (n=108) went through remanufacturing and resulted in the following:
    • Cancellation of products, 11.9% (n=46)
    • Within specification products, 10.9% (n=42)
  • Among the initial OOS products, 7.3% of products (n=28) were cancelled.
• Remanufacturing using material from patients with ≥4 prior LOT (n=199) resulted in the same or additional OOS results or cancellations as the first manufacturing OOS outcomes (67% of attempts, n=36).
  • Among initial OOS products, 35.7% of products (n=71) went through remanufacturing and resulted in:
    • Cancellation of products, 17.6% (n=35)
    • Within specification products, 12.1% (n=24)
• The cPDL during remanufacturing was higher when patients had fewer prior LOTs.

Recollection Outcomes

• Recollection of new apheresis material for remanufacturing was requested in 12% of initial OOS products due to insufficient remaining material or concerns regarding prior therapy effects, indicating a long-term impact of previous treatments and/or disease characteristics.
  • Remanufacturing using newly collected apheresis material (n=46) resulted in similar or worse outcomes in 31% of products, while 20% of products were cancelled.
  • The cancellation rate for remanufacturing in patients with ≥4 prior LOTs was 17.6% (n=35).

Clinical Outcomes of OOS vs In‑Specification CARVYKTI Production

Dingli et al (2025)^3,5 presented CARVYKTI outcomes for OOS products, comparing safety and efficacy with in-specification products.

Study Design/Methods

• Patients with MM treated with commercially approved CARVYKTI at Mayo Clinic between June 2021 and September 2024 were evaluated to compare outcomes of in-specification and OOS products.
  • Among the identified patients, 51 received in-specification product and 22 received an OOS product.

Results

Patient Characteristics

• The baseline demographic, clinical, and laboratory characteristics of these patients are presented in Table: Baseline Demographic, Clinical, and Laboratory Characteristics of the Patients Treated With CARVYKTI.
• There were no significant differences in the patient characteristics between the two cohorts.

Lymphocyte Kinetics After CARVYKTI Infusion

• There was significant difference in absolute lymphocyte count (ALC) and on various days early after CARVYKTI, and the recovery of the ALC was slower in patients who received an OOS product and the peak ALC achieved was also lower in patients with the OOS product; however, by day 15 after CARVYKTI, there appeared to be no difference between the two cohorts.

Efficacy

• At 3 months after CARVYKTI infusion:
  • None of the patients who had received an in-specification product had disease progression, and 77.5% achieved a very good partial response or better response.
  • A similar depth of response was observed in 66.7% of patients treated with an OOS product (P=0.5359, Fisher exact test), although 2 patients in the latter cohort experienced progression during the same period.
  • No significant difference was observed in the frequency of minimal residual disease-negative bone marrow results (P=0.4327).
• At median follow-up of 1.4 years after CARVYKTI infusion, 2 patients who received an OOS product and 4 patients who received an in-specification product experienced disease progression. Responses are summarized in Table: Efficacy Outcomes.

Safety

• The incidence of cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are presented in Table: CRS and ICANS in In-Specification vs OOS CARVYKTI Product.
  • Duration of ICANS was shorter in patients who received OOS product.

Cellular and Clinical Factors Affecting CAR T-cell Manufacturing in RRMM

Vucinic et al (2024)⁴ presented real-word data on factors influencing autologous lymphocyte collection and subsequent manufacturing of CARVYKTI or ide-cel in patients with RRMM focusing on cellular composition and T cell senescence.

Methods

• Cellular composition of 84 leukapheresis collections from 61 patients with RRMM intended for CARVYKTI (n=48) or ide-cel (n=36) manufacturing was retrospectively analyzed.
  • Target yield was defined as ≥1×10⁹ CD3+ cells.

Results

Manufacturing Outcomes

• CD3+ cell yields had a median of 31.5×10⁸ (range 4.9-318), with the target achieved in 73 of 84 collections.
• Overall, 74% of collections resulted in successful CAR-T production, 3% were terminated, and 23% produced OOS products.
• Reasons for OOS included insufficient CAR-T cell count (37%), insufficient proliferation (37%), insufficient interferon (IFN)-gamma potency assay (10%), or other causes (16%).
  • Manufacturing failure rates were similar (P=0.99) for ide-cel (9/27) and CARVYKTI (13/35).
  • In OOS productions, exposure to bispecific antibodies and cytomegalovirus replication prior to apheresis were more common, but the differences were not significant (P=0.13 and P=0.4).
• Manufacturing failures were not associated with sex, age, adverse cytogenetics or remission status prior to apheresis.
• Patients requiring ≥1 collection had a higher median number of prior LOTs (9 [range, 5-13] vs 7 [range, 1-14]; P=0.04).

Biomarker Association

• Patients who received OOS products, exhibited higher median CD3+ concentration in peripheral blood at apheresis compared with those who received successful product (770/μL [range, 82-4748] vs 462/μL [range, 71-3616]; P=0.002) and higher median CD3+ yields (39×10⁸ [range, 8-318] vs 29×10⁸ [range, 5-176]; P=0.04), respectively.
• The higher yields of senescent CD3+CD27-CD28- were observed to be risk factors for OOS. See Table: Median Yields of CD4+, CD8+, and Senescent T-Cell Subsets.
• A cutoff of ≥17.4×10⁸ CD3+CD27-CD28- cells (area under the curve=0.72) yielded a positive predictive value of 0.52 (95% CI, 0.40-0.86) and a negative predictive value of 0.84 (95% CI, 0.80-0.97) for manufacturing failure.

Efficacy

• Among patients who received OOS products (CARVYKTI [n=2], ide-cel [n=2]), (insufficient cell number: n=3, insufficient IFN-gamma potency: n=1), of which 3 achieved complete response and 1 had stable disease as best response after CARVYKTI cell infusion.
• After a median follow-up of 10 months, progression-free survival did not differ compared with patients who had successful first production (P=0.8).

Literature Search

A literature search of MEDLINE^®, Embase^®, BIOSIS Previews^®, and Derwent Drug File databases (and/or other resources, including internal/external databases) was conducted on 30 January 2026.