Survival Patterns of Non-Randomized Patients in the IELSG37 Study: Deauville Score 4 Is Not Necessarily Associated with Poor Outcome in Primary Mediastinal Lymphoma
Bibliographic record
Abstract
Background:Primary mediastinal B-cell lymphoma (PMBCL) is clinically and biologically distinct from other types of aggressive lymphoma, it is characterized by a bulky mediastinal mass and most commonly occurs in women between 30 and 40 years of age. Although consolidation radiotherapy (RT) has yielded extremely positive outcomes in PMBCL, it carries a higher risk of long-term complications. Consequently, there has been an ongoing debate regarding the necessity of RT in patients responding to frontline immunochemotherapy. Methods:In the IELSG37 study, PMBCL 530 patients were treated with rituximab- and doxorubicin-containing regimens; 98% had stage I-II. Their response was assessed by positron emission computed tomography (PET/CT); 268 achieved a complete metabolic remission (CMR) defined as Deauville score (DS) 1 to 3 according to the Lugano classification and were randomly allocated to observation or mediastinal RT. The primary analysis [Zucca et al. J Clin Oncol 2023; 41 (suppl 17): abs. LBA7505] has shown that patients in CMR can safely forgo RT. The 3-year overall survival (OS) rate was 99% irrespective of irradiation. RT provided only minimal benefit on progression-free survival (PFS), which was over 96% at 3 years in both arms. Here we present the outcome of the 262 patients who were not randomized and were managed according to the preference of their treating physician. Results:Median follow-up was 64 months (interquartile range [IQR], 49-69). Besides 230 patients failing frontline treatment (174 DS4 and 56 DS5), the non-randomized cohort also included 1 patient with DS2 (who refused randomization and received RT) and 31 patients with DS3 who were initially considered as partial responders before a protocol amendment that changed the definition of CMR from DS 1-2 to DS 1-3. Among these non-randomized patients with DS3, 27 had RT consolidation, 1 was only observed, 2 had salvage chemotherapy ± RT, and 1 had missing data due to consent withdrawal; their 3-year PFS was very close to the one of the randomized patients with DS3 (90 vs 92%, p=0.99). In the DS4 group, 12 patients were observed, 149 had only RT, and 13 had second-line chemotherapy, with (12) or without (1) autologous stem cell rescue, 8 of 13 also had RT. Among the patients with DS5, only 2 were observed, 32 had only RT, and 22 had second-line chemotherapy, with (15) or without (7) autologous stem cell rescue; 12 of them also had RT. The outcome of patients with DS4 was overlapping the one of the patients achieving a CMR, with 3-year PFS of 97.1% (95%CI 93-99) and 97.4% (95%CI 95-99), respectively, while patients with DS5 had a significantly poorer 3-year PFS (62.4%, 95%CI 48-74) [Figure1A]. The 3-year OS showed a similar trend (98% in patients with DS4, 99% in those randomized, and 78% in the DS5 group). Notably, compared to patients receiving consolidation RT alone or only observed, the residual lesions on post-immunochemotherapy PET/CT scans of patients with DS4 and DS5 who were treated more aggressively (with salvage chemotherapy ± autologous transplant and/or RT) displayed significantly higher tracer maximum standardized uptake value (SUVmax, median: 8.2, IQR 4.6-16.5 vs. 3.8, IQR 3.1-4.9; p<0.0001) and larger residual metabolic tumor volume (MTV, median: 18 ml, IQR 3-95 vs. 3.6, ml, IQR 1-10; p<0.0001). These patients had significantly worse outcomes compared to those who received consolidation RT or were observed only [Figure 1B]. Conclusions:Patients with DS4 had outcomes similar to those of patients achieving a CMR, indicating that this group may either include a significant number of “false-positive” cases with inflammatory uptake or that their residual disease is limited and can potentially be cured with consolidation RT. Patients with DS5 had a significantly poorer outcome, although nearly half of them did well with RT alone. It seems plausible that clinical condition of the patient and/or imaging features of the residual disease have influenced treatment choices in patients without CMR after frontline immunochemotherapy. Further studies are needed to investigate whether PET metrics and/or liquid biopsies can aid in personalizing treatment and identifying patients who could potentially avoid irradiation.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".