--- title: "CAR-T Cell Therapy as Second-Line Treatment for R/R Aggressive B-Cell Lymphoma" slug: cart_dlbcl_2l version: 1.1 timestamp: 2026-04-20T00:00:00Z date: 2026-04-20 specialty: Haematology (Lymphoma) canonical_url: https://mahmood726-cyber.github.io/rapidmeta-finerenone/protocols/cart_dlbcl_protocol_v1.1_2026-04-20.md app_url: https://mahmood726-cyber.github.io/rapidmeta-finerenone/CART_DLBCL_REVIEW.html license: MIT --- # Second-Line CAR-T vs SoC in Relapsed/Refractory Aggressive B-Cell Lymphoma ## A Living Systematic Review and Meta-Analysis Protocol **Protocol version:** 1.1 **Frozen:** 2026-04-20T00:00:00Z **Specialty:** Haematology (Lymphoma) --- ## 1. Review Title and Registration **Title:** Second-Line CD19-Directed CAR-T Cell Therapy vs Standard-of-Care Chemoimmunotherapy plus Autologous Stem-Cell Transplant in Relapsed or Refractory Aggressive B-Cell Lymphoma: A Living Systematic Review and Meta-Analysis of Pivotal Phase 3 RCTs **Registration:** Protocol frozen 2026-04-20T00:00:00Z. Pre-registration mechanism: **GitHub-canonical-URL freeze**. **Authors:** Mahmood Ahmad (corresponding). **Corresponding address:** drmahmoodclinic@pm.me (GMC 6071047), Royal Free Hospital and Barnet Hospital, London, UK. --- ## 2. PICO | Element | Specification | |---|---| | **Population** | Adults with aggressive B-cell lymphoma (DLBCL, HGBL, PMBCL) refractory to first-line chemoimmunotherapy or relapsed within 12 months | | **Intervention** | CD19-directed CAR-T cell therapy: axicabtagene ciloleucel, lisocabtagene maraleucel, or tisagenlecleucel | | **Comparator** | Salvage chemoimmunotherapy (2-3 cycles) + BEAM or BEAM-R conditioning + autologous stem-cell transplant for responders | | **Outcome (primary)** | Event-free survival (trial-specific composite of disease progression, death, new anti-lymphoma therapy, or stable disease at designated assessment) | --- ## 3. Eligibility Criteria ### Inclusion - RCT (parallel, open-label with IRC-adjudicated primary) - Phase III pivotal trials - R/R aggressive B-cell lymphoma (refractory to 1L or relapsed <=12 months) - CD19-directed CAR-T at licensed dose - SoC comparator = salvage + ASCT intent - EFS as primary endpoint ### Exclusion - Third-line or later CAR-T trials (separate review) - Follicular lymphoma CAR-T trials (ZUMA-5 — separate) - Mantle-cell lymphoma CAR-T trials (ZUMA-2, TRANSCEND-NHL-001 — separate) - CD19-BiTE / bispecific antibody trials (mosunetuzumab, glofitamab, epcoritamab) - CD22- or CD20-directed CAR-T (not licensed at freeze) --- ## 4. Information Sources and Search Strategy | Database | Query | |---|---| | ClinicalTrials.gov | `(axicabtagene OR lisocabtagene OR tisagenlecleucel) AND (DLBCL OR aggressive B-cell lymphoma) AND second-line` | | Europe PMC / PubMed | `(ZUMA-7 OR TRANSFORM OR BELINDA) AND randomized` | --- ## 5. Study Selection Two-reviewer adjudication with HMAC-signed seal; PRISMA 2020. --- ## 6. Data Extraction Study IDs, N randomized, refractory/relapsed strata, ORR, CR rate, manufacturing time, bridging therapy use, EFS events, cytokine release syndrome / neurotoxicity rates, RoB 2. --- ## 7. Risk of Bias Assessment Cochrane **RoB 2**. Design-based priors: - **All 3 trials:** D2 SOME CONCERNS (open-label by necessity — CAR-T single infusion vs multi-cycle chemo). - **BELINDA specifically:** D2 SOME CONCERNS is particularly salient — the 52-day median time from randomisation to tisa-cel infusion plus 26% bridging therapy use may have systematically disadvantaged the CAR-T arm, contributing to the negative primary result. - Other domains LOW for all three (IRC adjudication, stratified randomisation, SAP lock). Author sign-off required per the provisional-RoB banner. --- ## 8. Synthesis / Statistical Methods - **Primary pool:** DerSimonian-Laird random-effects IVW on log-HR scale for EFS. - **CI adjustment:** HKSJ with t-distribution df=k-1; floor `max(1, Q/(k-1))`. - **Prediction interval:** t-df=k-2 (Higgins 2009), k=3 so PI enabled. - **Subgroup / meta-regression:** By CAR-T product (axi-cel vs liso-cel vs tisa-cel), refractory vs early-relapse, manufacturing-time strata. - **Sensitivity:** Leave-one-out (critical given BELINDA's heterogeneous result); restrict to non-tisa-cel products. - **Heterogeneity:** Expected to be HIGH because BELINDA's HR 1.07 vs ZUMA-7 0.40 / TRANSFORM 0.35 is mechanistically distinct (bridging + manufacturing time). - **Zero-cell handling:** Conditional 0.5 correction. - **Publication bias:** Low power at k=3; contour-enhanced funnel reported. - **Trim-and-fill sensitivity:** Low-power caveat at k<10. - **Bayesian:** Grid-approximation with half-normal(0, 0.5) prior on tau. - **Browser-hosted WebR cross-validation (optional, user-triggered):** Analysis tab button loads WebR + `metafor` on first click. --- ## 9. Certainty of Evidence (GRADE) Standard GRADE domains with a **pre-specified indirectness downgrade (`serious`)** for the class-level pool because the three CAR-T products differ in manufacturing time, bridging-therapy permissiveness, and conditioning regimen. Product-specific inference is preferred over a single class pool. Inconsistency downgrade is also anticipated given the BELINDA null result. --- ## 10. Reporting and Dissemination - **Reporting guideline:** PRISMA 2020 - **Published benchmark:** No single external IPD MA combining ZUMA-7 + TRANSFORM + BELINDA at protocol freeze. Internal DL pool + heterogeneity narrative; agent-level results cited directly. --- ## 11. Living-MA Update Cadence - **Trigger:** New 2L CAR-T RCTs, long-term survival updates (ZUMA-7 5-year, TRANSFORM 3-year, BELINDA final), real-world comparative cohorts. - **Formal 3-monthly cadence:** Quarterly search + protocol-check. - **Change-of-estimate threshold:** > MCID or > half 95% CI width triggers v1.x amendment. --- ## 12-14 Aggregate-data only. No competing interests, no funding. --- ## Appendix A. AMSTAR-2 Critical-Domain Self-Assessment | # | Domain | Self-rating | Evidence / location | |---|---|---|---| | 2 (critical) | Protocol registered before data extraction | Yes (partial) | GitHub-canonical-URL freeze. | | 4 (critical) | Comprehensive search strategy | Yes | §4. | | 7 (critical) | List of excluded studies with reasons | Yes | Screening tab. | | 9 (critical) | RoB assessment | Yes | Authors have double-checked screening and data extraction against the record excerpts shown in the Extraction tab; the provisional AI-drafted RoB-2 ratings have been author-confirmed. Formal dual-assessor RoB-2 with inter-rater kappa remains a per-submission artefact (see Extraction-tab banner). | | 11 (critical) | Appropriate statistical methods | Yes | §8, pre-specified indirectness downgrade for class pool. | | 13 (critical) | RoB accounted for in interpretation | Partial | GRADE §9 + BELINDA manufacturing/bridging narrative. | | 15 (critical) | Publication bias assessment | Yes (k-appropriate) | §8 k=3 so formal tests low-power. | | 1 | PICO research question | Yes | §2. | --- ## Changelog - **v1.1** (2026-04-20) — Initial editor-review-revision release.