SOHO 2026: Disparities, Efficacy, and Long-Term Safety in B-Cell Lymphoma

The 2026 Society of Hematologic Oncology (SOHO) Annual Meeting highlighted several poster presentations across the aggressive B-cell lymphoma (ABCL) track addressing unresolved questions in the treatment and management of relapsed or refractory B-cell lymphoma, from national data characterizing sex-specific inpatient outcome disparities to emerging long-term safety data with established bispecific agents.

Here are some of the key takeaways from the ABCL poster sessions.

Women With Aggressive B-Cell NHL Experience Fewer Inpatient Complications But Markedly Higher Depression Rates Than Men

A retrospective cross-sectional analysis of the National Inpatient Sample (NIS) 2019-2020 presented by Aayushi Pareek, MD, a PGY-2 Internal Medicine resident from St. Bernards Medical Center in Jonesboro, Arkansas, examined sex-based differences in inpatient complications, resource utilization, and costs among patients hospitalized with aggressive B-cell non-Hodgkin lymphoma (NHL).¹

The final cohort included 40,532 discharges, of whom 15,791 (39%) were female and 24,741 (61%) were male. Multivariable generalized estimating equations logistic regression clustered by hospital was performed, adjusting for age, race and ethnicity, insurance type, income quartile, NHL subtype, hospital characteristics, Elixhauser comorbidity score, admission type, and calendar year, with false discovery rate correction applied across all outcomes.

At baseline, female patients were older (64.1 vs 61.8 years; P <.001), had higher DLBCL prevalence (81.0% vs 71.6%), higher Medicare enrollment (55.4% vs 48.6%), and significantly higher documented depression prevalence (17.5% vs 10.3%; standardized mean difference [SMD] = 0.21). Female sex was not independently associated with inpatient mortality (adjusted odds ratio [aOR], 0.93; 95% CI, 0.84-1.03; P = .169). However, women demonstrated significantly lower odds of acute kidney injury (aOR, 0.68; 95% CI, 0.64-0.72), sepsis (aOR, 0.89; 95% CI, 0.82-0.97), shock (aOR, 0.78; 95% CI, 0.66-0.92), and mechanical ventilation (aOR, 0.86; 95% CI, 0.77-0.96); all significant after false discovery rate correction. Women also incurred $827 lower median hospitalization costs ($14,434 vs $15,261; P <.001), representing an estimated $65 million in annual national savings.

“Women hospitalized with aggressive B-cell NHL experience fewer severe complications and lower costs despite equivalent mortality, suggesting sex-based dimorphism in acute physiological stress response,” Pareek wrote with coauthors in the poster.¹ “The higher depression burden in women is an underrecognized concern warranting psycho-oncologic screening.”

How Do Approved Third-Line Large B-Cell Lymphoma Therapies Compare on Efficacy and Safety?

Nain Tara, MD, an internal medicine resident from Guthrie Robert Packer Hospital in Sayre, Pennsylvania, presented a Bayesian network meta-analysis comparing complete response (CR) rates and safety across all currently approved third- or later-line therapies for relapsed/refractory large B-cell lymphoma (LBCL), including CAR-T cell products, bispecific antibodies, antibody drug conjugates, and other agents.²

The analysis identified 10 studies encompassing 1258 patients and 11 treatment nodes. Because 9 of the 10 studies lacked a control arm, an arm-based parametrization was used with synthetic bendamustine (Treanda) plus rituximab (Rituxan) comparators anchored to the phase 1b/2 GO29365 trial (NCT02257567) CR rate (17.5%). A Bayesian random-effects model confirmed convergence (Gelman-Rubin Ř max = 1.04), with substantial between-study heterogeneity noted (τ = 1.08).

CR rankings by SUCRA (surface under the cumulative ranking curve) placed axicabtagene ciloleucel (axi-cel; Yescarta) first (SUCRA 75.5%; P[rank 1] = 29.7%), followed by lisocabtagene maraleucel (liso-cel; Breyanzi; 72.5%), epcoritamab (Epkinly; 58.3%), tafasitamab (Monjuvi) plus lenalidomide (Revlimid; 58.1%), polatuzumab vedotin-piiq (Polivy) plus bendamustine/rituximab (pola-BR; 57.4%), glofitamab (Columvi; 56.1%), and mosunetuzumab (Lunsumio; 47.8%). Loncastuximab tesirine (Zynlonta) and selinexor (Xpovio) were included but SUCRA values were not individually reported for these agents. No pairwise comparison reached statistical significance given the indirect nature of the comparisons across predominantly single-arm data (axi-cel vs synthetic BR: OR, 6.09; 95% CrI, 0.16-207.7).

CAR-T CR rates were derived from manufactured-to-treat populations, with 22% to 31% of enrolled patients never receiving infusion, a factor that inflates observed CR rates relative to off-the-shelf agents. Pooled grade 3 or higher cytokine release syndrome (CRS) was 12% with CAR-T vs 3% with bispecific antibodies; grade 3 or higher neurotoxicity was 17% vs 2%, respectively. Among bispecific antibodies, epcoritamab (Epkinly) was ranked most favorably when accounting for both efficacy and safety.

“CAR-T therapies ranked highest for CR rate but with greater toxicity and selection bias,” Tara wrote with coauthors in the poster.² “Among bispecific antibodies, epcoritamab ranked most favorable for combined efficacy and safety. The near-disconnected, star-topology network with predominantly single-arm data means results carry very low certainty and warrant cautious interpretation.” The authors also noted that time-to-event analyses, including progression-free survival and overall survival, are needed to complement response-rate rankings and better differentiate these agents.

What Is the Real-World Risk of Second Primary Malignancies With CD20/CD3 Bispecific Antibodies?

Carmel Awadallah, MD, internal medicine resident from St John’s Episcopal Hospital in Far Rockaway, New York, and Muhammad Bilal Abid, MD, medical director of the Blood and Marrow Transplantation and Cellular Therapy program at Texas Tech University Health Sciences Center, presented a retrospective pharmacovigilance analysis characterizing the frequency, characteristics, and disproportionality safety signals of second primary malignancies (SPMs) associated with the 3 FDA-approved CD20/CD3 bispecific antibodies—mosunetuzumab, epcoritamab, and glofitamab—using real-world data from the FDA Adverse Event Reporting System (FAERS), spanning quarter 2 of 2022 through quarter 4 of 2025.³

Among 4,237 FAERS lymphoma reports featuring epcoritamab (n = 2368), glofitamab (n = 1138), and mosunetuzumab (n = 731), 29 unique SPM cases were adjudicated after deduplication (12 certain or probable; 17 uncertain), yielding a pooled SPM rate of 0.68%. Per-drug crude SPM rates were highest with glofitamab (1.14%; n = 13/1,138), followed by epcoritamab (0.51%; n = 12/2368) and mosunetuzumab (0.41%; n = 3/731); no disproportionality signal was detected at the individual-drug level. The underlying lymphoma was predominantly DLBCL (53.6%) and follicular lymphoma (17.9%); the median age of patients with an SPM was 70 years (range, 29-90), and 73.7% were aged 65 years or older.

Hematologic malignancies accounted for 42.9% of SPMs, primarily acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS; n = 7), consistent with known therapy-related myeloid neoplasm risk in a heavily pretreated cohort. Solid tumors comprised 39.3% of SPMs (skin, 14.3%; gastrointestinal, 14.3%; breast, 10.7%). The median time to SPM onset was 515 days (range, 1-1096). Case fatality among SPM cases was 21.4% (n = 6/28), highest with glofitamab (28.6%) and lowest with mosunetuzumab (0%).

At the class level, CD20/CD3 bispecific antibody SPM reporting was significantly lower than for other lymphoma drugs (reporting odds ratio [ROR], 0.47; 95% CI, 0.26-0.86; P = .009) and significantly lower than for CAR-T therapy (ROR, 0.11; 95% CI, 0.06-0.23; P <.001).

“SPM events associated with CD20/CD3 bispecific antibodies were rare, with no disproportionality signal detected at the class or individual-drug level and significantly lower SPM reporting vs CAR-T therapy,” Awadallah and Abid concluded in the poster.³ “These findings provide reassuring real-world safety data for the CD20/CD3 bispecific class; ongoing surveillance with longer follow-up is warranted as post-marketing exposure accumulates.”

References

1. Pareek A, Yeramalla SD, Nappier P, Durrani H. Sex-based disparities in resource utilization and complications in patients hospitalized with aggressive B-cell non-Hodgkin lymphoma: a national inpatient analysis 2019-2020. Presented at: 2026 Society of Hematologic Oncology (SOHO) Annual Meeting; September 9-12, 2026; Houston, TX.
2. Tara N, Zubair H, Joy J, Adhikari A, Tayyab H, Talamo G. Comparative efficacy and safety of approved therapies in ≥3L relapsed/refractory large B-cell lymphoma: a Bayesian network meta-analysis. Presented at: 2026 Society of Hematologic Oncology (SOHO) Annual Meeting; September 9-12, 2026; Houston, TX.
3. Awadallah C, Abid MB. Secondary primary malignancies associated with CD20×CD3 bispecific T-cell engaging antibodies in patients with B-cell non-Hodgkin’s lymphoma. Presented at: 2026 Society of Hematologic Oncology (SOHO) Annual Meeting; September 9-12, 2026; Houston, TX.