Comparing T-cell lymphoma treatments: what drug lists and response rates leave out

Two different recommendations often leave patients comparing the number of drugs, their price or response rates found online. The underlying disagreement may be elsewhere: the clinicians may be using different diagnoses, assessing different treatment stages or pursuing different immediate goals. Aligning those assumptions comes before ranking the regimens. Otherwise the comparison may put answers to different clinical questions in the same column.

Key takeaways

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

Ask each clinician to state the complete entity, disease extent, previous treatment and current response. Systemic ALCL, PTCL-NOS and a TFH-derived lymphoma should not be reduced to one general label. The 2025 ESMO–EHA guideline itself separates entities and disease settings; it does not provide a universal ranking of drugs for everyone with a T-cell lymphoma. d’Amore et al.: ESMO–EHA guideline for peripheral T- and NK-cell lymphomas, 2025
Overall response generally includes complete and partial responses under the chosen criteria. Duration of response concerns those who responded and does not represent everyone who started treatment. Progression-free survival and overall survival answer different questions again. Response frameworks such as Lugano help define assessments, but a reader still needs to know the actual study method and whether review was independent. Cheson et al.: Lugano classification for lymphoma evaluation and response
Create a short comparison covering the goal and evidence, administration setting, main hazards, monitoring, likely subsequent treatment and costs that remain unconfirmed. Use the same currency and service scope. One cycle's drug cost cannot fairly be compared with another option's complete hospital admission. Where no formal itemised quotation exists, mark the item as awaiting confirmation instead of filling the gap with an unsourced online total.

Quick answer

Full guide

Check that both recommendations address the same decision

Ask each clinician to state the complete entity, disease extent, previous treatment and current response. Systemic ALCL, PTCL-NOS and a TFH-derived lymphoma should not be reduced to one general label. The 2025 ESMO–EHA guideline itself separates entities and disease settings; it does not provide a universal ranking of drugs for everyone with a T-cell lymphoma. d’Amore et al.: ESMO–EHA guideline for peripheral T- and NK-cell lymphomas, 2025

Clarify the aim as well. Seeking a durable first remission, achieving control before a transplant, relieving pressure on an organ and reducing the burden of long-term treatment are different decisions. Rapid shrinkage can be valuable without establishing that a regimen is the best long-term option. An outpatient approach can be convenient without necessarily addressing an urgent organ threat adequately.

Frame the disagreement as a specific question: in this confirmed subtype and treatment setting, what comparative evidence supports adding or replacing this medicine? If the diagnoses differ, resolve the tissue review first. Statistical comparisons cannot provide a sound answer when the recommendations do not share the same diagnostic foundation.

BV-CHP versus CHOP depends on the studied population

ECHELON-2 directly randomised patients between BV-CHP and CHOP and its longer follow-up provides information about disease control and survival. The trial enrolled CD30-expressing PTCL, with systemic anaplastic large cell lymphoma forming the majority. Its findings therefore have a different degree of direct applicability to a systemic ALCL case than to a rare non-ALCL subgroup with few participants. Horwitz et al.: ECHELON-2 five-year results

Ask how many participants had the relevant entity, whether the subgroup has sufficiently robust evidence and why the proposed replacement is appropriate. Exploratory subgroup observations can inform discussion without amounting to an independently proven result for every subtype. Also establish exactly which components change. Adding every drug named in two regimen abbreviations is not a valid interpretation of a comparison.

Treatment in China additionally requires verification of the current domestic indication, prescribing basis and hospital supply. National Health Commission drug guidance distinguishes Chinese uses from indications listed for other jurisdictions. An overseas approval cannot settle the local prescribing question, and an applicable indication does not itself confirm that a particular hospital can provide the full course and monitoring. 中国国家卫生健康委：《新型抗肿瘤药物临床应用指导原则（2025年版）》，2026-01-26发布

Adding an active medicine does not guarantee a better combination

Ro-CHOP compared romidepsin plus CHOP with CHOP. Its final analysis did not establish a clear overall efficacy advantage across the studied first-line PTCL population; exploratory findings in a TFH-related subgroup require their own cautious interpretation. Activity of a drug by itself in relapsed disease cannot prove that adding it to initial chemotherapy will improve the overall result. Bachy et al.: Final analysis of the randomised Ro-CHOP trial, 2024

A combination also needs assessment of severe cytopenias, infection, discontinuation and the ability to complete treatment. A striking response in an individual story does not answer whether average benefit outweighs the added burden. If a non-standard combination is suggested, ask whether it is part of a trial or individualised treatment and why the uncertainty is considered reasonable in this case.

A study failing to meet its objective does not prove that the drug has no possible role in any setting. The conclusion belongs to a particular population, combination and purpose. Different partners, schedules or stages require their own evidence. Both enthusiastic and dismissive claims can become misleading when the original study's limits disappear.

Single-arm response rates cannot create a reliable league table

JACKPOT8 Part B studied golidocitinib in relapsed or refractory PTCL in a single-arm phase 2 setting. It describes responses and safety in its selected population, without randomising those participants against chidamide, another single agent or transplantation. Differences in prior treatment, subtype composition and eligibility can change the apparent response rate independently of the treatment being compared. Song et al.: JACKPOT8 Part B, phase 2 golidocitinib study

When two separate studies are placed side by side, check previous regimens, inclusion of refractory disease, response definitions, follow-up and discontinuations. Overall response in one study should not be compared with complete response in another as if they were identical outcomes. A clinician can still recommend an option in the absence of direct comparison, while acknowledging that those numbers cannot prove an absolute ranking.

NCI's PTCL summary organises several relapse treatments and drug classes; it is not an individual sequence of prescriptions. Previous resistance, infection history, neurological function and marrow recovery may narrow the useful options. Those features are more informative than marketing descriptions such as the latest generation of treatment. NCI PDQ: Peripheral T-Cell Non-Hodgkin Lymphoma Treatment

Response and survival describe different outcomes

Overall response generally includes complete and partial responses under the chosen criteria. Duration of response concerns those who responded and does not represent everyone who started treatment. Progression-free survival and overall survival answer different questions again. Response frameworks such as Lugano help define assessments, but a reader still needs to know the actual study method and whether review was independent. Cheson et al.: Lugano classification for lymphoma evaluation and response

A median is a position within a studied population, not a deadline assigned to a patient. A median not yet reached does not mean events will never occur; follow-up may simply be immature. If a study reports only early response, uncertainty about sustained control should be made explicit. Symptom relief, longer life and preserved daily function can all matter, but evidence for one outcome should not silently substitute for another.

Autologous and allogeneic transplantation have different trade-offs

The cell source and biological effects differ between autologous and allogeneic transplantation. Donor immunity may help control lymphoma but also brings risks such as graft-versus-host disease. The EBMT Handbook discusses selection by entity, response and patient circumstances. A donor transplant is not inherently an upgraded version suitable for every person considered for an autologous procedure. Hübel et al.: Other B- and T-Aggressive Lymphomas, EBMT Handbook 2024

A randomised study in younger patients with specified-risk PTCL compared autologous and allogeneic transplantation within first-line treatment. It illustrates the balance between relapse prevention and transplant-related hazards. Its eligibility and design do not establish that every newly diagnosed patient should receive an allogeneic transplant, nor can its conclusions be transferred without qualification to all relapsed cases or modern transplant approaches. Schmitz et al.: Randomised autologous versus allogeneic transplantation trial in PTCL, 2021

The 2026 EBMT registry analysis reports outcomes after allogeneic transplantation in major T-cell entities. A registry can describe real practice but remains subject to selection: people who reach transplantation have already passed several clinical filters. Comparing their unadjusted outcomes with those of people who did not undergo transplantation can mistake selection differences for the treatment's entire effect. EBMT Lymphoma Working Party: allogeneic transplantation for major T-cell lymphoma entities, 2026

Maintenance needs evidence for the maintenance question

JACKPOT26 reported golidocitinib maintenance after response to first-line therapy in 2026. This differs from treating active relapsed or refractory disease, and its participants were selected by prior response. Without a randomised comparator, later remission in the treated cohort does not establish that every patient benefits from maintenance rather than observation. Wei et al.: JACKPOT26 maintenance study, Blood Cancer Journal 2026

If maintenance is proposed, ask about the entity, current response, supporting evidence, intended purpose and criteria for stopping or reassessing it. Observation still needs follow-up; continued medication brings monitoring and adherence requirements. The practical comparison is not simply taking a tablet versus doing nothing. The team should explain how the individual's risk and preferences influence the choice.

Compare burdens using the same scope

Create a short comparison covering the goal and evidence, administration setting, main hazards, monitoring, likely subsequent treatment and costs that remain unconfirmed. Use the same currency and service scope. One cycle's drug cost cannot fairly be compared with another option's complete hospital admission. Where no formal itemised quotation exists, mark the item as awaiting confirmation instead of filling the gap with an unsourced online total.

Infection history and access to urgent care can affect feasibility. NCI's neutropenia information explains why infection may become an emergency during treatment. Being able to take a medicine at home does not eliminate the need for hospital support. Confirm blood monitoring and an out-of-hours contact pathway, particularly when the patient intends to return to another country. NCI: Infection and Neutropenia during Cancer Treatment

Ask the clinician what might be gained and what might be given up by choosing the alternative recommendation. The answer cannot guarantee an outcome, but it should identify the evidence and uncertainty behind the current preference. Patients do not have to perform their own statistical review to request an explanation that is specific, open to questions and compatible with their daily circumstances.