2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Monday, December 14, 2026
06:00 p.m. - 08:00 p.m.
MCCNO - Hall F
Technology and Procedures, Molecular Testing, Pathology, Treatment Considerations, Combination Therapy, Biological Therapies, Transplantation, (CAR)-T Cell Therapies, Non-Biological Therapies, Chemotherapy, Artificial Intelligence (AI), Machine Learning, Biological Processes, Genomics, Molecular Biology, Clinical Practice (Health Services and Quality), Diseases, Lymphoid Malignancies, Lymphomas, Non-Hodgkin Lymphoma, B Cell Lymphoma, Aggressive Lymphoma, Research, Translational Research, Clinical Research, Epidemiology, Health Outcomes Research, Real-World Evidence, Registries, Survivorship, Study Population, Human, Adult
PAPER

5402 Clinical impact of cell of origin determination by Hans IHC algorithm vs gene expression profiling: A prognostic non-inferiority meta-analysis

Ash A. Alizadeh, MD, PhD, Elias Campo, MD, PhD2, Jordan Scott Goldstein, MD, MSc3, Girish Venkataraman, MD4, Shady Gendy, PharmD, MBA5, Barbara Emoingt Furtado, MD, MBA6, Aniket Das7, Anirban Basu, BS, MS, PhD, FASA8
1 Stanford University School of Medicine, Palo Alto, California, United States
2 Institute of Biomedical Research August Pi Sunyer (IDIBAPS), University of Barcelona, Hospital Clinic of Barcelona, Barcelona, Spain
3 Stanford University, Palo Alto, California, United States
4 University of Chicago, Chicago, Illinois, United States
5 AstraZeneca, Gaithersburg, Maryland, United States
6 AstraZeneca, Gaithersburg, Maryland, United States
7 Indence Health, Kolkata, India
8 University of Washington, Seattle, Washington, United States

Background: In diffuse large B-cell lymphoma, not otherwise specified (DLBCL-NOS), cell of origin (COO) classification has established prognostic relevance and is recognized in international guidelines. While gene expression profiling (GEP) is commonly viewed as the analytical benchmark, platform heterogeneity, tissue requirements and operational complexity have resulted in uneven real-world adoption. A recent meta-analysis (Alizadeh et al., 2026 [ENHANSEMENT]) found that COO classification by the widely used Hans immunohistochemistry (IHC) algorithm (CD10/BCL6/MUM1) met a prespecified non-inferiority (NI) margin of 10% for diagnostic accuracy compared to GEP. Here, we assess the clinical impact of this non-inferiority on prognostic performance. We extended the non-inferiority framework to progression-free survival (PFS) to test whether Hans IHC diagnostic non-inferiority to GEP translates into clinically meaningful prognostic non-inferiority in COO-based classification in DLBCL-NOS.

Methods: We used the PRISMA-driven evidence base established for the prior meta-analysis ENHANSEMENT, designed to mitigate key limitations of prior IHC-GEP comparisons, including differential verification bias and inclusion of entities beyond DLBCL-NOS. Eligible studies reported PFS by COO subgroup: germinal center B-cell-like DLBCL (GCB) and non-GCB (defined in GEP as activated B-cell-like (ABC) plus Unclassified cases). The primary endpoint was the hazard ratio (HR) for PFS in non-GCB versus GCB. The primary analysis was restricted to studies reporting GCB versus non-GCB PFS contrasts for both Hans IHC- and GEP-defined COO within the same patient cohort, referred to as same-patient studies. Overall survival (OS) was evaluated as a secondary analysis. Sensitivity analyses evaluated differences in PFS between both tests using restricted mean survival time (RMST), and a broader study set combining all PFS-reporting GEP and IHC studies from different cohorts, referred to as pooled-patient studies. A Bayesian hierarchical model used diagnostic accuracy estimates from the original study to compare each method’s observed HR (non-GCB relative to GCB) with the latent HR expected under a hypothetical perfect test. Hans IHC was considered non-inferior to GEP if the upper bound of the 95% credible interval (CrI) for the ratio of predicted hazard ratios (GEP over Hans IHC) did not exceed the same 10% NI margin of the original study.

Results:

In ENHANSEMENT, 19 studies (N=2,247 tumor samples) showed Hans IHC to be non-inferior to pooled GEP for diagnostic accuracy, with estimated accuracy of 90% (95% CrI, 87-93) for Hans IHC and 95% (93-96) for GEP; the 4.5% difference (1-8) met the prespecified 10% NI margin.

In this analysis, 7 studies reported PFS (N=1,155 patients), including pooled-patient studies. Among these, 3 same-patient studies (N=235) contributed to the primary analysis. After accounting for misclassification by both methods, the estimated HR for PFS in non-GCB relative to GCB disease was 1.37 (0.96-1.89) for Hans IHC and 1.42 (0.95-2.01) for GEP. The GEP-to-Hans IHC HR ratio met the 10% NI margin at 1.03 (0.99-1.09), with the upper bound remaining below 1.10. Findings remained consistent when expanding to the full study pool to include all 7 evaluable PFS studies across 1,155 patients.

For the secondary OS analysis, 4 same-patient studies (N=384) were included. The GEP-to-Hans IHC HR ratio was 1.06 (1.01-1.13), and the upper bound exceeded the 10% NI margin.

In absolute terms, using RMST, Hans IHC was associated with an estimated 12-day difference versus GEP in COO-based PFS prognostic performance (1-28 days) and a 14-day difference in OS prognostic performance (2-29 days). These differences were well below the shortest reported RMST treatment benefit margins leading to FDA oncology drug approvals (2.5%-ile of ~80 days for PFS and ~40 days for OS) across 94 oncology randomized Ph2/3 trials (Rahmadian et al., 2020), suggesting limited impact on obscuring such margins.

Conclusions: Together, this meta-analysis now establishes that Hans IHC is non-inferior to GEP both diagnostically and prognostically based on PFS outcomes for COO classification, supporting its use as a practical and generally reliable method for COO classification in DLBCL-NOS in routine practice.
Ash A. Alizadeh, MD, PhD: Resero - Current equity holder in private company; CiberMed - Current equity holder in private company; Celgene - Research Funding; Bristol Myers Squibb - Research Funding; Roche - Patents & Royalties; Foresight Diagnostics - Patents & Royalties; Janssen - Consultancy (Includes expert testimony); Roche - Consultancy (Includes expert testimony); Chugai - Consultancy (Includes expert testimony); Adaptive - Consultancy (Includes expert testimony); Foresight Diagnostics - Consultancy (Includes expert testimony); Capstan - Consultancy (Includes expert testimony); Gilead - Consultancy (Includes expert testimony); Genentech - Consultancy (Includes expert testimony); Pharmacyclics - Consultancy (Includes expert testimony); Natera - Consultancy (Includes expert testimony); Arima - Consultancy (Includes expert testimony); ADCT - Consultancy (Includes expert testimony); CiberMed - Consultancy (Includes expert testimony); Celgene - Consultancy (Includes expert testimony); Resero - Consultancy (Includes expert testimony); AstraZeneca - Consultancy (Includes expert testimony)