Highlights
In patients with Hodgkin lymphoma enrolled in the GHSG HD21 trial, circulating tumor DNA minimal residual disease measured after two treatment cycles (MRD-2) was strongly associated with subsequent relapse, progression, or death.
Using the validated LymphoVista sequencing assay, MRD-2 positivity identified a biologically high-risk subgroup with markedly inferior progression-free survival compared with MRD-2-negative patients.
When integrated with positron emission tomography after two cycles (PET-2), MRD-2 appeared to refine early risk classification, separating very low-risk from very high-risk patients.
The study supports ctDNA-based MRD as a potential treatment-guidance tool in Hodgkin lymphoma, but prospective response-adapted intervention studies remain necessary before routine clinical adoption.
Background
For most patients with classical Hodgkin lymphoma, especially in contemporary cooperative group trials, cure is an achievable goal. The next frontier is more difficult: identifying who can safely receive less treatment, who should receive intensified or biologically adapted therapy earlier, and how to reduce the short- and long-term burden of treatment. This issue is especially important in Hodgkin lymphoma because many patients are young and treatment-related late effects, including infertility, cardiopulmonary injury, and second cancers, can shape survivorship for decades.
Interim fluorodeoxyglucose positron emission tomography (PET), commonly after two cycles of therapy, has transformed response assessment in Hodgkin lymphoma. PET-adapted strategies have already influenced treatment selection in both early-stage and advanced-stage disease. Yet PET is not perfect. It reflects metabolic activity rather than direct molecular persistence of malignant clones, can be affected by inflammatory changes, and may not fully capture residual systemic disease burden. A complementary biomarker with higher molecular sensitivity could therefore improve early prognostication.
Circulating tumor DNA (ctDNA) has emerged as a promising tool across lymphoid malignancies. In Hodgkin lymphoma, the biological rationale is particularly compelling. The malignant Hodgkin/Reed-Sternberg cells constitute only a minority of the tumor mass within an inflammatory microenvironment, making tissue-based molecular monitoring challenging. By contrast, plasma ctDNA can potentially provide a real-time, whole-body molecular readout of disease burden and treatment response. However, ctDNA MRD studies in Hodgkin lymphoma have varied widely in assay design, pre-analytical methods, time points, and thresholds for MRD negativity, limiting cross-study comparability and clinical implementation.
The current Blood report by Heger and colleagues addresses this translational gap by applying LymphoVista, a validated ctDNA sequencing platform, to samples from the German Hodgkin Study Group HD21 trial. The central question was clinically important and practical: does MRD assessed after two cycles of front-line treatment provide early and robust prognostic information in advanced Hodgkin lymphoma?
Study Design
Parent trial context
This translational analysis was embedded in the GHSG HD21 trial, a major randomized study in advanced-stage Hodgkin lymphoma comparing BrECADD with escalated BEACOPP (eBEACOPP). The HD21 platform is well suited for biomarker work because it enrolled a large, uniformly treated population with standardized imaging and clinical follow-up.
Biomarker strategy
The investigators used LymphoVista, a validated ctDNA sequencing assay developed for lymphoma genotyping and MRD monitoring. Plasma samples were analyzed after two cycles of treatment, defining the key biomarker time point as MRD-2. The article specifically addresses outcome associations of this early on-treatment molecular response assessment.
Design features
The analysis used a case-cohort design. This is an efficient epidemiologic sampling framework in which a randomly selected subcohort from the parent trial is studied alongside all or most outcome events of interest. Such an approach allows biomarker evaluation without requiring assay testing of the entire trial population while preserving the ability to estimate event associations. Because this design alters the underlying sampling fraction, the investigators also used inverse probability weighting to estimate outcome metrics aligned with the full reference set.
Endpoints
The principal clinical endpoint was progression-free survival (PFS), defined in the trial context as freedom from relapse, progression, or death. The investigators examined the association between MRD-2 status and subsequent PFS, and they also explored the combined prognostic value of MRD-2 and interim PET-2.
Key Results
MRD-2 positivity marked substantially inferior prognosis
The most clinically striking finding was the clear separation in outcome between MRD-2-positive and MRD-2-negative patients. In the analyzed cohort, 4-year PFS was 36.7% among patients with positive MRD-2 compared with 82.2% among those with negative MRD-2. The hazard ratio for relapse, progression, or death was 5.3, with a 95% confidence interval of 2.0 to 13.8, and the association was statistically significant at P = .0008.
This effect size is notable. A hazard ratio above 5 suggests that persistence of detectable ctDNA after only two cycles identifies a population with markedly increased risk of treatment failure. In practical terms, MRD-2 appears not merely to be prognostic in a modest way, but to define a biologically meaningful subgroup with resistant or incompletely responsive disease.
Weighted analysis suggested stronger absolute outcomes in the full reference set
Because the analysis used a case-cohort design, the authors performed inverse probability weighting to account for the number of events in the full reference set. After this adjustment, the estimated 4-year PFS rates were 72.2% for MRD-2-positive patients and 95.3% for MRD-2-negative patients.
This weighted result is important for interpretation. It suggests that while the unweighted case-cohort sample highlights the strong discriminatory value of the biomarker, the expected absolute PFS rates in the broader trial population were higher overall, particularly among MRD-negative patients. Even so, the risk separation remained clinically meaningful. A weighted 4-year PFS of 95.3% in MRD-2-negative patients indicates the possibility of identifying a particularly favorable subgroup early during therapy.
MRD-2 and PET-2 together refined risk stratification
The study also found that combining MRD-2 with PET-2 improved early prognostic resolution. According to the abstract, this joint approach identified patients at very low risk and patients at very high risk of relapse, progression, or death. This is biologically plausible and clinically attractive: PET captures metabolic response, whereas ctDNA captures molecular persistence of disease. The two modalities likely represent overlapping but nonidentical dimensions of treatment response.
Although the abstract does not provide the full subgroup counts or exact hazard estimates for each combined PET/MRD category, the clinical message is clear. Concordant PET-2-negative and MRD-2-negative findings likely define a group with excellent disease control, whereas persistent positivity by both methods likely flags major residual risk. Discordant cases may be especially informative for future treatment adaptation studies, because molecular and imaging biomarkers may not become negative at exactly the same pace.
Why this matters clinically
The value of an early biomarker in Hodgkin lymphoma depends on actionability. The present data do not by themselves prove that changing therapy based on MRD-2 improves outcomes. However, they establish the key prerequisite: early ctDNA status carries clinically relevant information. This opens the door to future interventional trials testing whether MRD-negative patients can safely receive less treatment and whether MRD-positive patients should receive earlier escalation, novel agents, or alternative regimens.
Expert Commentary
Strengths of the study
The major strength of this work is that it evaluates ctDNA MRD within a rigorously conducted prospective phase 3 trial infrastructure rather than in a small single-center observational cohort. That setting improves confidence in sample quality, treatment homogeneity, endpoint adjudication, and follow-up.
A second strength is use of a validated assay. One of the recurring limitations in ctDNA MRD literature is heterogeneity of platforms and thresholds. By using LymphoVista, the investigators move the field closer to a reproducible and clinically translatable framework.
Third, the focus on a clearly defined time point, after two cycles of therapy, is clinically sensible. This aligns with established interim PET practice and offers a natural decision node for future response-adapted studies.
Interpretive cautions
Several limitations deserve attention. First, this was a prognostic biomarker analysis, not a biomarker-directed treatment trial. Therefore, the study cannot establish that modifying therapy on the basis of MRD-2 improves survival or reduces toxicity.
Second, case-cohort designs are methodologically sound but analytically more complex than testing the full trial population. The inverse probability weighting approach appropriately addresses this, yet readers should distinguish between the raw observed event rates in the biomarker sample and the weighted estimates intended to reflect the larger reference set.
Third, the abstract does not provide full technical details on assay sensitivity, failure rates, or the proportion of patients with unevaluable samples. These issues matter in real-world implementation because even highly promising biomarkers can lose utility if plasma collection, sequencing depth, or baseline trackability are inconsistent.
Fourth, generalizability beyond the HD21 treatment context remains to be proven. The parent trial involved BrECADD and eBEACOPP, intensive regimens used primarily in advanced-stage disease and most often in specialized centers. Whether identical MRD dynamics and thresholds apply to ABVD-like regimens, brentuximab vedotin-containing programs, PD-1-based approaches, or older and frailer populations is unknown.
Biological plausibility
The observed association between persistent ctDNA after two cycles and inferior PFS is biologically coherent. Effective frontline therapy should rapidly reduce tumor burden and thereby lower circulating tumor-derived DNA fragments. If ctDNA remains detectable despite initial treatment, it likely reflects either a larger residual disease reservoir, less chemosensitive biology, sanctuary sites not fully captured by blood kinetics, or a combination of these factors. The strength of the prognostic signal in this study supports the idea that molecular persistence is not an incidental laboratory finding but a surrogate of treatment-resistant disease biology.
Relationship to the broader field
Previous work has already suggested that ctDNA levels at baseline and on treatment correlate with outcomes in Hodgkin lymphoma. What distinguishes the current study is its use of a validated assay and its anchoring in a large cooperative group trial. This helps move ctDNA MRD from exploratory promise toward a more standardized biomarker conversation.
The next step for the field is not simply more retrospective confirmation. It is prospective utility testing: for example, randomizing MRD-2-positive patients to standard management versus early incorporation of non-cross-resistant or immune-based therapy, or testing whether MRD-2-negative and PET-2-negative patients can complete treatment with reduced cumulative toxicity.
Clinical Implications
If these findings are confirmed prospectively, MRD-2 could help address three major goals in Hodgkin lymphoma care.
First, it could improve de-escalation. Patients who are both MRD-2-negative and PET-2-negative may have such a low relapse risk that treatment intensity or duration could potentially be reduced, limiting infertility risk, hematologic toxicity, infection, and long-term organ damage.
Second, it could support earlier escalation or biologic adaptation. Patients with persistent ctDNA positivity after two cycles may represent a group in whom waiting for later clinical failure is suboptimal. Such patients could become candidates for intensified imaging surveillance, earlier biopsy of suspicious residual lesions, or future trials incorporating antibody-drug conjugates, checkpoint inhibitors, or other targeted approaches.
Third, ctDNA-based monitoring may improve affordability and global practicality if eventually standardized. PET remains resource-intensive and variably accessible worldwide. A blood-based assay will not replace imaging in the near term, but a validated molecular test could become a valuable complement, particularly if future studies define settings in which it reduces repeated imaging or identifies patients who most need advanced scans.
Funding and ClinicalTrials.gov
The abstract provided does not specify funding details for this translational analysis. The parent study was conducted within the German Hodgkin Study Group framework. Readers should consult the full Blood article for complete funding disclosures and conflicts of interest.
The GHSG HD21 trial is registered at ClinicalTrials.gov as NCT02661503.
Conclusion
This study provides some of the strongest evidence to date that early ctDNA-based minimal residual disease assessment can meaningfully stratify risk in advanced Hodgkin lymphoma. In the GHSG HD21 trial setting, MRD positivity after two treatment cycles was associated with a substantially higher risk of relapse, progression, or death, while MRD negativity identified a group with very favorable progression-free survival. The combination of MRD-2 and PET-2 appears especially promising for defining very low-risk and very high-risk populations early in treatment.
For clinicians, the immediate message is prognostic rather than practice-changing. MRD-2 by LymphoVista is not yet a standard decision tool, but it is rapidly becoming a credible candidate biomarker for response-adapted therapy. For investigators and trialists, the implications are more direct: the field now has a validated molecular approach that appears sufficiently informative to justify prospective intervention studies.
In Hodgkin lymphoma, where cure must increasingly be balanced against lifelong treatment burden, that is an important advance.
References
Heger JM, Mattlener J, Kaul H, Ferdinandus J, Schneider J, Schleifenbaum JK, Schneider G, Schaub V, Hänel M, Hellmuth JC, Dierlamm J, Martin S, Mathas S, Meissner J, Pegtel DM, Zijlstra JM, Ossowski A, Becker K, Hallek M, von Tresckow B, Borchmann P, Borchmann S. MRD-2 in the GHSG HD21 trial assessed by a validated circulating tumor DNA sequencing assay. Blood. 2026 May 7;147(19):2194-2202. PMID: 41662627.
Borchmann P, Goergen H, Kobe C, et al. PET-guided treatment in patients with advanced-stage Hodgkin’s lymphoma (HD18): final results of an open-label, international, randomised phase 3 trial by the German Hodgkin Study Group. Lancet. 2018;390(10114):2790-2802.
Fuchs M, Plütschow A, Goergen H, et al. Positron emission tomography-guided treatment in early-stage favorable Hodgkin lymphoma: final results of the international, randomized phase III HD16 trial by the German Hodgkin Study Group. J Clin Oncol. 2019;37(31):2835-2845.
Borchmann P, Eichenauer DA, Plütschow A, et al. PET-guided omission of radiotherapy in early-stage unfavorable Hodgkin lymphoma: final results of the GHSG HD17 trial. J Clin Oncol. 2021;39(24):2628-2636.