New DLBCL treatments and clinical trials: evidence that can change a decision

New treatments have changed parts of first-line and relapsed DLBCL care, but new can mean several things: an approved regimen in one jurisdiction, a combination supported by a randomized study whose local access still needs checking, or an experimental approach whose safety and efficacy remain under investigation. These should not be merged into a shopping list. Start with the exact diagnosis and treatment stage. [S1,S3]

Key takeaways

These excerpts come from the original article. Read the full sections below for context.

POLARIX compared Pola-R-CHP with R-CHOP in previously untreated intermediate- or high-risk DLBCL. The five-year report published in 2025 gave progression-free survival rates of 64.9% and 59.1% in the global intention-to-treat population, with a hazard ratio of 0.77. Overall survival was not statistically significantly different at that analysis. [S24] These findings should not be restated as a fixed personal increase in life expectancy.
A trial record helps check the question, intervention, endpoint, locations and criteria. For example, EPCORE DLBCL-1 is NCT04628494 and studies epcoritamab against investigator-choice chemotherapy in relapsed or refractory DLBCL. [S27] This identifies the study; it is not a promise that a China site currently has a place.
A usable recommendation states the best standard option, why a trial or new combination is being considered, which criteria are met, what evidence is missing and what happens if enrollment or treatment fails. Unknowns should be specific enough to compare, even when no one can resolve them immediately.

Quick answer

Full guide

The examples below use identifiable primary studies and regulatory documents. Publication date differs from data cutoff, and foreign approval does not establish Chinese approval, reimbursement or stock. A receiving hospital must check the current product and eligibility requirements before offering an individual plan.

Interpret the longer follow-up for Pola-R-CHP

POLARIX compared Pola-R-CHP with R-CHOP in previously untreated intermediate- or high-risk DLBCL. The five-year report published in 2025 gave progression-free survival rates of 64.9% and 59.1% in the global intention-to-treat population, with a hazard ratio of 0.77. Overall survival was not statistically significantly different at that analysis. [S24] These findings should not be restated as a fixed personal increase in life expectancy.

The FDA first-line indication has defined adult pathology and IPI criteria. [S4] Ask how closely you resemble the trial population and how neuropathy or other risks will be managed. Localized or distinct-subtype disease needs its own assessment. A good response to ongoing R-CHOP is not a reason for self-directed switching after reading a newer paper.

Understand the bispecific-antibody category

Epcoritamab and glofitamab use a dual-target design to engage T cells, with different routes, step-up schedules and treatment duration. Their US indications concern specified relapsed or refractory large B-cell lymphoma populations and previous therapy. [S6,S7] Approval of a combination in another lymphoma does not establish the same combination's DLBCL indication.

Ask where the first doses are observed, whether treatment has a fixed endpoint, whether later doses can be delivered at home and how infection is prevented. Lack of individualized cell manufacturing does not remove cytokine release syndrome or neurological risk. Positioning before or after CAR T cells requires a clinical and evidence-based discussion.

Later safety updates also matter: the January 2026 glofitamab warning identifies HLH, with deaths reported. Its presentation may resemble CRS while requiring different management. Ask whether this update is incorporated into monitoring and consent. [S33]

Keep STARGLO within its actual clinical setting

STARGLO randomized previously treated, transplant-ineligible relapsed or refractory DLBCL patients to glofitamab-GemOx or rituximab-GemOx and reported an overall-survival advantage for the former. [S25] A three-year follow-up published in 2026 continued to report favorable outcomes. [S26] The trial did not directly compare the combination with CAR T cells.

Ask whether the prior-treatment history, transplant eligibility and organ status fit that population, and whether the regimen is locally available. Publication does not establish acceptance of international patients at a Chinese hospital. Step-up dosing, the chemotherapy component and the complete schedule need to be prescribed together; purchasing one component is not a treatment plan.

Recognize the precise population for a newer brentuximab combination

In 2025, the FDA approved brentuximab vedotin with lenalidomide and rituximab for specified adults after at least two lines who were ineligible for autologous transplantation or CAR T cells. ECHELON-3 supported benefit in survival and other endpoints, with risks including cytopenias, infection and neuropathy. [S29]

This is not evidence for adding brentuximab to every newly diagnosed DLBCL regimen. Nor does transplant ineligibility remove the need to assess other suitable options. Record full generic names: brentuximab and polatuzumab are different medicines with different evidence and combinations. Pharmacy verification is useful when names sound similar in translation.

Check whether earlier regulatory claims remain current

Older lists may still present selinexor as having a US DLBCL indication. An FDA letter dated April 30, 2026 confirms removal of that indication from the Xpovio labeling and discusses its voluntary withdrawal. [S28] A historical approval announcement should not be treated as today's status.

Regulatory simplification also differs from risk elimination. In 2025, the FDA removed certain autologous CAR T REMS requirements, while important risks such as cytokine release syndrome and neurological toxicity remained in product information. [S19] These are US changes, not automatic changes to Chinese labels, monitoring or hospital procedures.

Ask what a CAR T research design actually changes

Research may address manufacturing, different or multiple targets, earlier treatment or relapse after cell therapy. The patient should know how the investigational product differs from established products, whether collection is autologous, what human data exist and what happens if manufacture fails or disease deteriorates. [S11]

Laboratory killing does not establish clinical benefit. Patients previously treated with CAR T cells need assessment of targets, persistent cytopenias, infection and organ effects. NCI's discussion of secondary T-cell malignancies also shows why long-term safety observation remains relevant. An early remission does not answer every later safety question. [S23]

Use a registry identifier to verify the study

A trial record helps check the question, intervention, endpoint, locations and criteria. For example, EPCORE DLBCL-1 is NCT04628494 and studies epcoritamab against investigator-choice chemotherapy in relapsed or refractory DLBCL. [S27] This identifies the study; it is not a promise that a China site currently has a place.

Registry status can lag behind local activity. A study recruiting somewhere does not mean every site is recruiting or accepting international participants. Before travel, obtain confirmation from the actual center about the cohort and preliminary documents. The presence of DLBCL in a title does not establish eligibility for every subtype or previous treatment history.

Distinguish trial phase and endpoint

Early studies may primarily identify dose and toxicity, whereas later randomized trials are better suited to comparing clinical outcomes against existing treatment. Overall response, complete response, progression-free survival and overall survival are different measures. Short-term activity cannot replace durable benefit. Subgroup findings also require attention to whether they were planned and adequately sized.

Ask about the comparator, independent assessment and outcomes that are still immature. A higher response rate in a different patient group does not necessarily outweigh a lower rate elsewhere. Consent should explain what is known and unknown rather than display only the most hopeful result. [S16,S17]

Review eligibility before organizing an overseas stay

Pathology review, previous lines, antigen or molecular results, blood counts, organ function, infection, central nervous system disease and prior transplantation or CAR T cells may all matter. A study can require a fresh biopsy, measurable disease, a washout period and sufficient follow-up access. The research team must review these conditions individually. [S17]

Do not stop steroids, antimicrobial therapy or other necessary medicines to pursue possible eligibility. If treatment is urgent, the usual and research teams should coordinate bridging and its effect on entry. Screening can fail, so a conventional alternative should be planned before the patient travels rather than discovered afterward.

Include safety and withdrawal arrangements in the decision

Know the overnight contact, emergency destination, caregiver requirements and proximity rules. Investigational immune or targeted treatment can still cause low counts, infection and organ injury. Fever of 38°C or above, marked breathlessness, altered thinking or persistent bleeding needs prompt assessment. [S8]

Consent should describe the right to withdraw, ongoing clinical care after stopping the study medicine and necessary safety follow-up. Withdrawal should not be presented as requiring abandonment of all treatment. The purpose of extra blood draws, biopsies and sample storage also needs a clear explanation. [S16]

Confirm time and financial responsibility in China

Separate screening, enrollment, dosing and follow-up. A screening visit may not lead to treatment that day, and changing health can alter eligibility. Ask about length of stay, assessments that cannot be done abroad, repeat travel and the transition to usual care if the study ends early. [S17]

In Chinese yuan, list routine and research-only tests, medicines, admission, complications, transport, accommodation and the caregiver. Identify who pays each item: sponsor, hospital, insurer or patient. Free study medicine does not establish free overall care. No individual quotation or support policy from an open center was verified here, so a total or reimbursement promise would be unsupported. [S10]

Preserve the evidence behind the proposal

A candidate-study sheet can record the identifier, contact, disease-stage requirements, unresolved criteria, main risks, standard alternative and unknown costs. Keep the original paper, regulatory page and consent version. Short summaries can lose distinctions between medicines, indications and statistical endpoints after repeated retelling.

Continue updating administration and response records after treatment starts so the home clinician understands the relationship between research and ordinary care. [S12] If a new result changes the plan, ask for the evidence and timetable. The useful reason to choose a new treatment is that it addresses the present clinical problem, not that it received recent publicity.

Leave with an answer that the family can act on

A usable recommendation states the best standard option, why a trial or new combination is being considered, which criteria are met, what evidence is missing and what happens if enrollment or treatment fails. Unknowns should be specific enough to compare, even when no one can resolve them immediately.

Patients can reasonably choose established care without research participation. Ability to complete monitoring, tolerance for uncertainty and willingness to remain away from home are legitimate factors. Involving the home doctor and arranging records transfer before buying travel supports a more continuous course of care.