Design: Prognostic analysis within phase 3 randomised trials (GHSG HD18, HD21)
Population: Patients with advanced-stage classic Hodgkin lymphoma at interim PET-2 (primary cohort 639)
Primary outcome: Progression-free survival from PET-2
Effect: DS5 vs DS1–3 HR 3.00 (95% CI 1.25 to 7.23); DS5a vs DS1–4 HR 2.57–5.47 across cohorts

The Lugano Imaging Committee recently split Deauville 5 into 5a (uptake more than twice liver, no new lesions) and 5b (new lesions). This analysis tested the split in advanced-stage classic Hodgkin lymphoma at PET after two cycles, using the German Hodgkin Study Group HD18 and HD21 randomised trials.

DS5a was uncommon — 4 to 6% across cohorts. In the uniformly treated primary cohort, DS5 carried inferior progression-free survival compared with DS1–3 (HR 3.00, 95% CI 1.25 to 7.23), and DS5a remained adverse against DS1–4 in every sensitivity cohort (HR 2.57 to 5.47). Under PET-adapted treatment, the new DS4 did not consistently separate from DS1–3.

This is the first validation of the refined score in randomised trial populations. It means a DS5 without new lesions is not a lesser category: it identifies a small high-risk group that haematologists will want flagged.

Adopt the refined Deauville scale in interim PET reports for Hodgkin lymphoma.
State DS5a or DS5b explicitly, not DS5 alone.
Give the reference liver SUV and the lesion SUVmax used for the call.
Discuss DS5a cases at the lymphoma MDT.

Why it matters

A score change on the report now carries validated prognostic weight for the haematologist deciding what comes next.

The statistics, in plain English

Hazard ratios of 2.6 to 5.5 mean progression was several times more frequent with DS5a. With only 29 to 67 DS5a patients per cohort, the intervals are wide, and overall survival could not be judged because deaths were few.