EORTC Cutaneous Lymphoma Tumour Group RT Recommendations
TL;DRExpert-opinion dose recommendations consolidating reduced-dose RT (4-12 Gy) across cutaneous lymphoma subtypes; no randomised trials underpin any of it.
The operative number for an RT reader is the dose floor, not the ceiling: 8-12 Gy in two or three fractions holds ≥92% 1-yr local control in MF, while 4 Gy underperforms there but suffices for CD4+ small/medium T-LPD (100% remission, no relapses). That split is the prescribing decision, and it argues against one ultra-low-dose default across subtypes.
In a patient with a symptomatic MF plaque or an indolent pcMZL/pcFCL lesion, this supports prescribing 8-12 Gy rather than a 30-40 Gy course; it does not extend to advanced MF, Sézary syndrome, or DLBCL leg type, where the authors concede combined-modality room for improvement.
The prescribing decision is where the floor sits by subtype: 8-12 Gy in two or three fractions holds ≥92% 1-yr local control in MF, while 4 Gy underperforms there yet suffices for CD4+ small/medium T-LPD. Pair a 4 Gy start in indolent B-cell disease with a four-month response assessment and escalation to 24 Gy.
11 details 1 trial watching
Expert opinion from the EORTC cutaneous lymphoma tumour group, tabulating published retrospective and small prospective series (Tables 1 and 2) into a dose algorithm (Figs 2A/2B). No pooled estimate, no protocol-registered synthesis.
Covers MF, Sézary syndrome, pcALCL, CD4+ small/medium T-LPD, pcMZL, pcFCL, and DLBCL leg type. Explicitly excludes the aggressive CTCL variants (subcutaneous panniculitis-like, gamma/delta, CD8+ epidermotropic, NK/T), where the authors say low-dose RT has a limited role.
MF plaques/tumours 8-12 Gy in two or three fractions; low-dose TSEBT 8-12 Gy for palliation and up to 24 Gy pre-autologous transplant; pcALCL few relapses at 20 Gy, palliative 2×4 Gy; CD4+ T-LPD 4 Gy in two fractions; DLBCL leg type consolidation reduced to 30 Gy, with 20 Gy reported post-systemic. Modalities named: electrons, photons, kilovoltage X-ray, brachytherapy.
The efficacy signal is uniformly high ORR across the tabulated series, but the discriminating result is dose-dependent local control: ≥92% 1-yr local control after low-dose MF RT versus an inferior rate at 4 Gy, and 28% vs 5% local relapse for 4 Gy vs 8-50 Gy in the ILROG registry series (p < 0.001).
Grade 3/4 toxicity is absent across most tabulated low-dose series. The one clear dose-toxicity signal in MF local RT is 27% grade 3/4 following 12 Gy versus 0% after 4-8 Gy; a prospective DLBCL leg-type cohort reported 14%.
The reference frame is the 30-40 Gy conventional standard that governed cutaneous lymphoma until roughly two decades ago. This document formalises the retreat from it, but does so on a base the authors concede is retrospective, so it codifies practice already in motion rather than establishing it.
The dose tables mix single-lesion and per-patient denominators and span decades of technique, so a 4 Gy series and a 40 Gy series are not comparing like populations. Several tabulated rows report dose comparisons as n.s. in cohorts far too small to exclude a real difference, which is not the same as equivalence.
The unresolved question is not whether reduced dose works but where its floor sits, and the answer looks subtype-specific rather than universal. The authors' own response-adapted proposal (escalate to a cumulative 24 Gy for residual disease or failure after 4 Gy) concedes that 4 Gy alone is a starting position, not a definitive prescription.
- Dose floor for ultra-low-dose RT in indolent cutaneous B-cell lymphoma n=52 · primary completion 2025-12 · 4 Gy/2 fx in early-stage PCBCL, n=52
- Whether low-dose TSEBT plus immunotherapy prolongs remission in advanced MF/SS
- RT dose after systemic therapy in DLBCL leg type
📚 Sources · 📄 1 paper
The longer read
The clinically load-bearing claim in this document is narrower than its title suggests. It is not that reduced-dose radiotherapy works in cutaneous lymphoma, which the field largely accepts, but that the dose floor differs by subtype in a way a single ultra-low-dose default would obscure. Read that way, the recommendations divide cleanly. For CD4+ small/medium-sized T-cell lymphoproliferative disorder, 4 Gy in two fractions produced complete remission with no local relapses across the two small series tabulated, and the biology, an indolent process that in some hands is managed by excision alone, makes that credible. For mycosis fungoides the same dose is explicitly flagged as giving inferior local control, with 8-12 Gy in two or three fractions holding a 1-year local control rate at or above 92%. A reader who takes away 'go low' without taking away 'how low depends on histology' has drawn the wrong conclusion from the same tables.
The indolent B-cell recommendation is the least settled part and the authors do not hide it. Two bodies of retrospective evidence point in opposite directions: several small series report comparable response and in-field recurrence between ultra-low-dose and higher-dose schedules, while a registry series with longer follow-up found 4 Gy gave inferior response rates and substantially more in-field relapse, 28% versus 5%. Follow-up duration is the most plausible reconciliation, since in-field failure after a low dose in an indolent lymphoma is a late event, and a series that stops watching early will not see it. That asymmetry should move confidence toward the longer-followed result rather than treating the two as an even split, and it is the reasoning behind the proposed response-adapted escalation to a cumulative 24 Gy for residual disease or failure at four months. That algorithm is a sensible hedge, but it is worth naming what it concedes: 4 Gy is being positioned as a first attempt with a planned salvage path, not as a definitive dose.
The toxicity argument for going low is weaker than it first appears, because it rests on an evidence base with almost no grade 3/4 events anywhere. When nearly every tabulated series reports no severe toxicity at doses from 4 Gy to 40 Gy, dose reduction is not principally buying a reduction in severe harm. Its real returns are elsewhere and are still worth having: fewer hospital visits, retreatability of a relapsing field, lower financial toxicity, and preserved options in a disease where patients accumulate irradiated skin over years. Those are legitimate reasons to prescribe less, but they are convenience and durability-of-options arguments, not safety arguments, and conflating the two would let a reader accept a local control decrement they did not intend to accept.
The structural limitation is that the tables cannot be read as a dose-response experiment. Denominators alternate between patients and lesions, series span decades of technique, and the recurring 'n.s.' annotations on dose comparisons come from cohorts far too small to exclude clinically meaningful differences. Absence of a detected difference in a 21-patient series is not evidence of equivalence. The authors' closing call for a prospective randomised trial with translational correlates is therefore the most honest sentence in the paper: for a disease this rare, an expert dose algorithm is a reasonable interim instrument, but it should be held as a starting prescription that a documented four-month response assessment can revise, not as a settled standard.