Large‐cell transformation of mycosis fungoides: Patterns of care and patient outcomes

Summary

Large‐cell transformation of mycosis fungoides (LCTMF) is rare, histologically distinct, with an aggressive clinical course; yet is not recognised as an independent entity in classification systems nor in staging systems for mycosis fungoides and Sézary syndrome (MF/SS). Herein, the patterns of care and survival outcomes for patients with LCTMF are described, with prognosis compared to published data of non‐transformed MF/SS. Eligibility required clinicopathological diagnosis of LCTMF (1/1/1990–31/10/2021), managed at Peter MacCallum Cancer Centre. Eighty‐three patients were eligible. Median follow‐up was 8.0 years. At the time of LCTMF, 36% had early‐stage MF (IA‐IIA), 76% had cutaneous‐only LCTMF. The most common first‐line treatments were localised radiotherapy (48%) and multiagent chemotherapy (23%). Median overall survival (OS) from LCTMF diagnosis was 3.5 years (95% [confidence interval] CI: 2.2–8.2). Three prognostic groups of LCTMF were identified: unifocal cutaneous only, multifocal cutaneous only and extracutaneous (median OS: 4.6, 2.5 and 1.1 years, respectively; p = 0.005). Unfavourable prognostic factors were advanced age and extracutaneous LCTMF. In conclusion, treatment pathways for patients with LCTMF were varied, and prognosis was poor, despite >1/3 having early‐stage MF. However, differences in prognosis were suggested, with unifocal cutaneous LCTMF associated with greater OS. Given prognostic differences from MF/SS, consideration to include LCTMF in staging systems is warranted.

Large‐cell transformation of mycosis fungoides is rare with poor prognosis. Presenting over 20 years of data, this retrospective study of 83 patients demonstrated a median overall survival (mOS) of 3.5 years. However, prognostic differences in three clinical subgroups were suggested: unifocal, multifocal, extra‐cutaneous (mOS: 4.6, 2.5, 1.1 years, respectively).

INTRODUCTION

Large‐cell transformation of mycosis fungoides (LCTMF) is a rare phenomenon that is presently not recognised within existing staging systems for mycosis fungoides/Sézary syndrome (MF/SS) ¹ nor as a separate entity within the World Health Organization (WHO) classification of primary cutaneous T‐cell lymphomas (CTCL). ² , ³ , ⁴ Yet, characteristically, LCTMF has a more locally aggressive and rapidly progressing clinical course than classical MF, and is associated with poor median survival despite the use of more intensive therapy, ranging from 12 months to 5 years. ⁵ , ⁶ , ⁷ , ⁸ , ⁹ Typically LCTMF presents as rapidly enlarging plaque(s)/tumour(s), with potential for multisite cutaneous or extracutaneous progression. Retrospective series have described LCTMF occurring more frequently in patients with advanced stage MF/SS (up to 67% in stage IV, compared to 1.4% in stage IA‐IIA disease). ⁵ , ⁶ , ⁷ , ⁸ LCTMF is defined histologically as the presence of large, atypical T cells or clusters of large cells with nuclei that are >4 times the normal size, which can be CD30 negative or positive. ¹⁰ , ¹¹ To date, predictive risk factors and molecular mechanisms for the development of LCTMF remain largely unknown. ¹² , ¹³ , ¹⁴ Due to disease rarity, there is a paucity of clinical data specific to patients with LCTMF comparing the efficacy of treatments and survival outcomes. Furthermore, mortality data for those with LCTMF are seldom reported separately within clinical trials of MF/SS, and on occasions, patients with LCTMF are excluded from trial entry. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹

This study aims to describe the patterns of care and outcomes for patients with LCTMF, presenting over 20 years of data on demographics, disease characteristics, treatment received and survival outcomes. Additionally, as a post hoc exploratory analysis, we analyse survival outcomes across three clinically distinct groups: unifocal cutaneous‐only disease, multifocal cutaneous‐only disease and extracutaneous disease. We also discuss and compare the prognosis of patients with LCTMF, relative to that of the broader population with MF/SS as reported in the published literature. ⁵ , ²⁰ , ²¹ , ²²

METHODS

Eligibility

Ethical approval was obtained from the Human Research Ethics Committee at Peter MacCallum Cancer Centre for this retrospective, single‐centre, observational study. Eligibility required a clinicopathological diagnosis of LCTMF, with histological confirmation, between 1 January 1990 and 1 October 2021. Central review of pathology was routinely performed. Demographic, clinicopathological and treatment data were collected from electronic medical records, with a cut‐off date of 15 June 2022. Patients with non‐primary cutaneous lymphomas with secondary skin involvement were excluded.

As per published definitions, LCTMF was defined histologically as the presence of >25% large, atypical T cells or clusters of large cells with nuclei that are >4 times the normal size. ⁸ , ²³ Staging of MF/SS used the tumour, node, metastases, blood (TNMB) staging system and International Society for Cutaneous Lymphomas (ISCL)/European Organization of Research and Treatment of Cancer (EORTC) classification ¹ (Table S1). Early‐stage disease was defined as stage IA–IIA; advanced stage as IIB–IV.

For cutaneous lesions, ‘patch’ was defined as lesions with minimal elevation or induration, and ‘plaque’ was defined as lesions with notable elevation or induration, as per ISCL/EORTC definitions. ¹ ‘Tumour’ indicated solid or nodular lesions at least 1 cm in diameter, with evidence of depth or vertical growth. ¹ ‘Unifocal’ referred to the presence of a single LCTMF lesion whereas ‘multifocal’ indicated multiple lesions. Folliculotropism was defined as MF with a predominantly folliculotropic pattern, typically sparing the interfollicular epidermis, with or without adnexal involvement beyond hair follicles, as per the WHO definition. ²⁴

Treatment categories

Treatments were classified into skin‐directed therapies, systemic therapies and best supportive care. Skin‐directed therapies included localised radiotherapy (RT) and phototherapy. Systemic therapies included multiagent chemotherapies (combination of at least two drugs), single‐agent chemotherapies, histone deacetylase inhibitors (HDACi), monoclonal antibody therapies, interferon alpha, haematopoietic stem cell transplantation, immunomodulatory drugs and extracorporeal photopheresis (ECP) (Tables S2 and S3). Topical steroid use was not grouped separately as this was assumed to be used concurrently by most patients during their treatment pathway.

Prognostic factors

The following parameters at diagnosis of LCTMF were analysed for their prognostic significance: age; stage; extent of cutaneous disease (unifocal vs. multifocal); presence of extracutaneous disease (nodal involvement, visceral disease, bone marrow or blood involvement); presence of folliculotropic variant and CD30‐positive status. CD30 expression was defined as positive if staining was present in >10% atypical lymphocytes, consistent with the recent phase III ALCANZA trial in an effort to standardise reporting. ²⁵

Statistical analysis

Overall survival (OS) was calculated from time of diagnosis of LCTMF to death from any cause, with patients censored at their last follow‐up or the close‐out date (15 June 2022), whichever occurred first. Median follow‐up time was estimated using the reverse Kaplan–Meier method. Time‐to‐event analyses for OS were performed using the Kaplan–Meier and competing risk methods. The competing risk analysis estimates the cumulative incidence of the probability of death from LCTMF and other causes and is interpreted as cause‐specific, lost life‐years. ²⁶

Cox proportional hazards regression model was used to estimate hazard ratios (HRs) for prognostic variables, with 95% confidence intervals (CIs). Age was initially modelled on a continuum, but diagnostics indicated a non‐linear association with the hazard of death up to age 50, after which the relationship became approximately linear. Due to the small number of patients aged <50 years, a binary variable for age >55 years was used in regression analyses, satisfying the proportional hazards assumption.

RESULTS

Patient demographics

Eighty‐three patients were eligible with confirmed histological diagnosis of LCTMF between 1 January 1990 and 1 October 2021. 58 (70%) patients were >55 years at diagnosis of LCTMF, 53 (64%) were men (Table 1). The median time from diagnosis of classical MF/SS to diagnosis of LCTMF was 2.1 (range: 0–27.2) years. Of the 83 patients, 64 (77%) were diagnosed with LCTMF de novo, while 19 (23%) had LCTMF diagnosed after an earlier diagnosis of classical MF/SS. Among the total cohort, six patients were diagnosed with SS: four patients had existing SS at the time of LCTMF diagnosis, one developed LCTMF prior to diagnosis with secondary SS and one patient was diagnosed synchronously with LCTMF and SS.

TABLE 1.

Demographics and disease characteristics of patients with large‐cell transformation of mycosis fungoides (LCTMF).

Characteristic	n = 83 (%)
Sex, n (%)
Male	53 (63.9)
Female	30 (36.1)
Age at diagnosis of LCTMF, years, median (range)	67.7 (37.3–86.5)
LCT confirmed at time of first diagnosis with MF, n (%)
Yes	64 (77.1)
No	19 (22.9)
Distribution of LCTMF, n (%)
Cutaneous only, unifocal	32 (38.6)
Cutaneous only, multifocal	31 (37.3)
Extracutaneous involvement	15 (18.1)
Cutaneous and extracutaneous disease (nodal/visceral organ/bone marrow/blood)	10
Extracutaneous disease only (nodal)	5
Unknown	5 (6.0)
Clinical stage of MF a at time of diagnosis of LCTMF, n (%)
IA	10 (12.0)
IB	15 (18.1)
IIA	5 (6.0)
IIB	27 (32.5)
IIIA	4 (4.8)
IVA1	4 (4.8)
IVA2	2 (2.4)
IVB	6 (7.2)
Not documented	10 (12.0)

Abbreviations: LCT, large‐cell transformation; MF, mycosis fungoides.

^a Staging according to ISCL/EORTC Classification. ¹

Disease characteristics

At the time of LCTMF diagnosis, 30 (36%) patients had early‐stage MF (patch/plaque, stage IA–IIA), 43 (52%) had advanced stage MF (stage IIB–IV) and MF staging was insufficiently documented in 10 (12%) patients. Cutaneous‐only LCTMF was diagnosed in 63 (76%) patients (32 unifocal and 31 multifocal). Extracutaneous LCTMF was identified in 15 (18%) patients at first presentation: nodal involvement in 13, visceral disease in 5, bone marrow infiltration in 2, and blood involvement in 1. In 5 (6%) patients, the extent of LCTMF could not be definitively confirmed retrospectively.

Pathological features

All patients diagnosed with LCTMF had ≥1 biopsy demonstrating histopathological confirmation of LCTMF. On skin immunohistochemistry, CD30 expression was measured in 81 (98%) patients: with positivity in 47 (57%) patients (positivity defined as staining present in >10% atypical lymphocytes). Of the 83 patients, 13 (16%) patients were reported to have folliculotropic MF.

First‐line therapy

Following diagnosis with LCTMF, there was heterogeneity in the first‐line management. First‐line therapies included skin‐directed therapies (local radiotherapy, phototherapy) and/or systemic therapies, including multiagent chemotherapy (such as cyclophosphamide, vincristine and doxorubicin), single‐agent chemotherapy (such as gemcitabine), interferon alpha (including pegylated Interferon), monoclonal antibody‐based therapies (such as brentuximab vedotin) and HDACi (including vorinostat, panobinostat, romidepsin) (Tables S2 and S3).

Overall, local radiotherapy was the most common first‐line therapy delivered in 40 (48%) patients, followed by multiagent chemotherapy in 19 (23%) (Table 2). Local radiotherapy was consistently the most commonly used first‐line therapy for patients with unifocal cutaneous‐only LCTMF (78%), multifocal cutaneous‐only LCTMF (29%) and extracutaneous LCTMF (40%).

TABLE 2.

Summary of treatment received in patients with large‐cell transformation of mycosis fungoides (LCTMF).

First‐line therapy after diagnosis of LCTMF, n (%)	Total (n = 83)
Skin‐directed therapy
Local radiotherapy	40 (48.2)
Phototherapy	2 (2.4)
Systemic therapy
Multi agent chemotherapy	19 (22.9)
Single agent chemotherapy	8 (9.6)
Interferon	7 (8.4)
Monoclonal antibody‐based therapy	3 (3.6)
Histone deacetylase (HDAC) inhibitor	2 (2.4)
Best supportive care (BSC)
BSC	2 (2.4)

Therapies used by patients in the treatment of LCTMF and MF/SS, across all lines, n (%)
Skin‐directed therapy
Any radiotherapy	66 (79.5)
Localised radiotherapy	60 (72.3)
Total skin electron beam therapy (TSEBT)	17 (20.5)
TSEBT as conditioning prior to haematopoietic stem cell transplantation	7 (8.4)
Phototherapy	5 (6.0)
Systemic therapy
Any systemic therapy	70 (84.3)
Multiagent chemotherapy	38 (45.8)
Single agent chemotherapy	43 (51.8)
HDAC inhibitor	25 (30.1)
Monoclonal antibody‐based therapy	19 (22.9)
Interferon alpha	18 (21.7)
Autologous haematopoietic stem cell transplantation	13 (15.7)
Immunomodulatory drugs	4 (4.8)
Extracorporeal photophoresis (ECP)	2 (2.4)
Allogeneic haematopoietic stem cell transplantation	0

Abbreviation: MF/SS, mycosis fungoides and Sezary syndrome.

All lines of therapy

Median follow‐up was 8.0 (95% CI: 6.1–11.0) years. Considering all lines of therapy received after LCTMF diagnosis, 70 (84%) patients received ≥1 line of systemic therapy during their treatment journey (median, 2 lines of systemic therapy; range, 0–11). Across all lines of therapy, local radiotherapy was received by 60 (72%) patients, multiagent chemotherapy by 38 (46%), single‐agent chemotherapy by 43 (52%), HDACi by 25 (30%), monoclonal antibody‐based therapy by 19 (23%) (including brentuximab vedotin in 7 (8%)), interferon by 18 (22%) and 13 (11%) underwent an autologous stem cell transplant (AuSCT) (no patients underwent allogeneic stem cell transplant). 17 (20%) patients received total skin electron beam therapy (TSEB), with 7 (8%) of those patients receiving it as preconditioning for an AuSCT (Table 2).

Overall survival, causes of death and probability of death from disease

From time of diagnosis of LCTMF, median OS for the whole cohort was 3.5 (95% CI: 2.2–8.2) years. At 5 and 10 years, OS rates were 41% (95% CI: 32–54) and 31% (95% CI: 21–46) respectively (Figure 1A). Wide CIs suggested a range of prognostic outcomes within the cohort, prompting further investigation in a post hoc analysis of survival per clinical subgroup. Differences were observed in survival outcomes between the three clinical groups: median OS for unifocal cutaneous‐only LCTMF, multifocal cutaneous‐only LCTMF and extracutaneous LCTMF were 4.6 (95% CI: 3.5–5.6) years, 2.5 (95% CI: 1.3–3.7) years and 1.1 (95% CI: 0–2.2) years respectively (Figure 2). On log‐rank test, the results indicated a statistically significant difference in OS between these three clinical groups (p = 0.005).

FIGURE 1.

Overall survival and mortality for patients with large‐cell transformation of mycosis fungoides (LCTMF) (n = 83): (A) Kaplan–Meier estimate of overall survival with 95% confidence intervals; (B) cumulative incidence of death.

FIGURE 2.

Overall survival of patients with large‐cell transformation of mycosis fungoides (LCTMF) in three clinical groups (p = 0.005). [Colour figure can be viewed at wileyonlinelibrary.com]

At the time of analysis, 54 patients were deceased: 40 deaths were from MF/SS‐related causes (tumour‐stage (T3) or visceral (M1) disease refractory to therapy), three from treatment‐related causes (death due to sepsis while receiving systemic treatment for multirelapsed MF/LCTMF), six from unrelated causes/competing comorbidities (including acute myocardial infarction, COVID‐19 infection and progressive neurodegenerative disease) and five unknown.

Of the 40 patients who died from MF/SS, 16 patients had early‐stage MF at the time of initial LCTMF detection and 20 had advanced stage MF/SS (with unknown stage in 4 patients). The probability of death from MF/SS increased rapidly for approximately 4 years after LCTMF detection and then slowed (Figure 1B). On average, patients lost 5 years of life in the first 10 years of follow‐up. Of those 5 years of life lost, 4.1 years were attributable to death from LCTMF and 0.9 years from other causes. The 3‐, 5‐ and 10‐year probabilities of death from LCTMF were 38% (95% CI: 28–49), 48% (95% CI: 37–59) and 52% (95% CI: 41–64) respectively. For unrelated/other causes, they were 6% (95% CI: 1–11), 10% (95% CI: 4–18) and 16% (95% CI: 6–26) respectively. After 6.9 years, the probability of an LCTMF death was 50%.

Predictors of outcome

On univariate analysis, factors associated with poorer OS included: presence of extracutaneous disease at LCTMF diagnosis (HR 1.92 (CI: 0.97–3.80)), age >55 years (HR 1.58 (CI: 0.84–2.96)), advanced stage of MF at diagnosis of LCTMF (HR 1.56 (CI: 0.84–2.87)), and multifocal cutaneous‐only involvement of LCTMF (HR 1.22 (CI: 0.67–2.23)) (Table 3). CD30 expression and folliculotropic variant were associated with improved OS (HR: 0.80 (CI: 0.46–1.39) and HR: 0.64 (CI: 0.27–1.51) respectively). However, the wide 95% CIs indicated uncertainty for all variables.

TABLE 3.

Predictors of outcome in patients with large‐cell transformation of mycosis fungoides (LCTMF) (n = 70).

Variable at diagnosis of LCTMF	Univariate analysis		Multivariate analysis
	HR	95% CI a	HR	95% CI a
Age
Age ≤ 55 years	1.00 (ref)		1.00 (ref)
Age >55 years	1.58	0.84–2.96	2.18	1.05–4.54
Presence of extracutaneous disease
No	1.00 (ref)		1.00 (ref)
Yes	1.92	0.97–3.80	2.37	1.07–5.24
Extent of skin involvement
Localised	1.00 (ref)		1.00 (ref)
Multifocal	1.22	0.67–2.23	1.04	0.55–1.98
Advanced stage of MF at diagnosis
Early stage (IA–IIA)	1.00 (ref)
Advanced stage (IB–IV)	1.56	0.84–2.87
CD30‐positive expression (>10%)
No	1.00 (ref)		1.00 (ref)
Yes	0.80	0.46–1.39	0.83	0.43–1.59
Folliculotropic variant
No	1.00 (ref)
Yes	0.64	0.27–1.51

Abbreviations: CI, confidence interval; HR, hazard ratio; MF, mycosis fungoides.

^a 95% confidence intervals for hazard ratio derived from multivariate Cox proportional hazards models.

The multivariable model included four variables, with data available from 70 patients and 43 events. Advanced age (>55 years) and extracutaneous disease at LCTMF diagnosis were both found to be independent adverse prognostic factors, with HR of 2.18 (95% CI: 1.05–4.54) and 2.37 (95% CI: 1.07–5.24) respectively (Table 3).

DISCUSSION

Overall, LCTMF is associated with a poor prognosis, with median OS of 3.5 years in this cohort. Unique to the literature, this study reveals variation in survival outcomes for patients with LCTMF, with three distinct clinical groups suggested: unifocal cutaneous‐only LCTMF was associated with a significantly more favourable prognosis, with a median OS of 4.6 years. In comparison, those with multifocal cutaneous‐only disease had a median OS of 2.5 years, while patients with extracutaneous disease had a significantly lower median OS of 1.1 years.

Identification of reliable prognostic factors is critical for predicting prognosis and clinical decision‐making. Despite typically long survivals reported in patients with MF, stratification by clinical stage demonstrates significant prognostic impact. For early‐stage (I–IIA) MF, median OS rates of 18–35 years have been observed ²⁰ , ²² , ²⁷ , ²⁸ , ²⁹ ; whereas in advanced stage (IIB–IV) MF/SS, median OS rates are as short as 1–6 years. ⁵ , ²⁰ , ²⁷ , ²⁸ , ²⁹ Although LCTMF has consistently been identified as an independent adverse prognostic factor, ⁷ , ¹⁰ , ²⁰ , ³⁰ , ³¹ it is not presently incorporated into the staging systems for MF/SS. ¹ , ³² , ³³ With median OS of 3.5 years, the poor prognosis of patients with LCTMF as seen in this study and others ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁵ , ¹⁶ , ²⁰ is comparable to the published survival outcomes of those with advanced stages of non‐transformed MF/SS ⁵ , ²⁰ , ²⁸ , ³⁴ , ³⁵ , ³⁶ (Table 4; Table S4). These findings suggest that LCTMF may represent an important consideration for prognostication; further evaluation is warranted to determine its potential inclusion in staging systems for MF/SS.

TABLE 4.

Comparison survival outcomes of large‐cell transformation of mycosis fungoides (LCTMF), all stages of mycosis fungoides and Sézary syndrome (MF/SS) and advanced stage MF/SS.

Study first author	No. of patients	Stage of disease	Number of patients with LCTMF, N (%)	Median OS, years	5‐year OS (%)	10‐year OS (%)
Large‐cell transformation of MF/SS
Current study (2025)	83	IA–IVB with LCTMF	83 (100)	3.5	41	31
All stages MF/SS
Agar et al. 20 (validation for EORTC staging)	1502	IA–IVB	70 (4.7)	18.3	75	62
Kim et al. 34	525	IA–IVB	Not recorded	11.4	68	53
Talpur et al. 28	1263	IA–IV	109 (8.6)	24.4	—	—
Advanced stage MF/SS
Alberti‐Violetti et al. 35	168	IIB–IV	43 (25.6)	2.5	31	—
Arulogun et al. 5	92	IIB–IVB	19 (20.7)	5.0	49 (predicted)	32 (predicted)
Campbell et al. 31	178	IVA with SS	35 (19.7)	5.3	53	—
Scarisbrick et al. 36	1275	IIB–IV	215 (16.9)	5.3	52	—

Despite distinct histopathological and clinical differences between non‐transformed MF and LCTMF, LCTMF is not presently recognised as a distinct diagnostic entity within the current WHO classification of CTCL. ³² MF is defined as an ‘uncontrolled proliferation of mature, skin‐homing cutaneous T‐lymphocyte antigen (CTLA)‐expressing, CD4+, CD45RO+ helper T cells’. ¹² Pathologically, there is proliferation of small cerebriform lymphocytes forming a dense infiltrate in the upper dermis. ¹¹ LCTMF occurs when these small neoplastic lymphocytes undergo molecular and genetic changes to produce clonally identical large cells within the tumour infiltrate, developing the morphologic appearance of a large cell lymphoma. ¹³ High mutational heterogeneity has been recognised with recurrent activating RAS mutations (Kirsten rat sarcoma viral oncogene homolog (KRAS) and neuroblastoma rat sarcoma viral oncogene homolog (NRAS)) being present. ³⁷ Generally, transformation involves decreased expression of pan T‐cell antigens (CD2, CD3, CD5). ²³ Importantly, LCTMF may be CD30 receptor positive or negative. The distinction of CD30+ LCTMF from CD30+ lymphoproliferative disorders (primary cutaneous anaplastic large‐cell lymphoma (pcALCL) and lymphomatoid papulosis (LyP)) is challenging but critical, with LyP being a benign condition associated with excellent outcomes. ¹¹ , ³⁸ , ³⁹ , ⁴⁰ In our cohort, only 7 of 47 patients with CD30+ LCTMF received brentuximab vedotin, limiting ability for meaningful survival analysis. However, brentuximab vedotin has shown significant improvement in median progression‐free survival (PFS) in patients with CD30+ MF: in post hoc subset analyses of the phase III ALCANZA trial (35% LCTMF in each arm), median PFS was 15.5 months in patients receiving brentuximab vedotin versus 2.8 months for physician's choice. ²⁵ , ⁴¹ Given this efficacy, brentuximab vedotin may be considered in the treatment of patients with CD30‐positive LCTMF.

In this study, a clinically significant proportion of patients had early‐stage MF at the time of first detection with LCTMF, contrasting with previous reports where LCTMF was described almost exclusively in advanced stage MF. ⁵ , ⁷ , ⁸ , ¹⁰ , ¹⁶ , ²⁰ , ⁴² In our cohort of 83 patients with LCTMF, 36% were first diagnosed with LCTMF while with early‐stage (SI–IIA) MF, compared to only 1.4% of patients in earlier data. ⁵ Additionally, the rate of de novo LCTMF diagnosis (77%) was also higher than previously reported in smaller studies (range: 32%–43%). ⁵ , ¹⁵ , ²³ We speculate that increased detection of LCTMF in MF early‐stage disease may be due to heightened vigilance or ‘diagnostic scrutiny’, leading to greater biopsy rates. These findings underscore the need for clinical suspicion of LCTMF across all stage groups.

In 2013, the Cutaneous Lymphoma International Prognostic Index (CLIPi) for MF/SS was proposed, based on a retrospective cohort of 1502 patients, to identify those at higher risk for poorer survival outcomes. ²⁰ , ⁴³ This study revealed older age, male sex, presence of plaques, folliculotropism and nodal involvement to be significant prognostic factors in early‐stage disease. In late‐stage disease, in addition to age and male sex, N2/N3 stage, B1/B2 stage and M1 visceral involvement were also identified as key prognostic factors. Our study of patients with LCTMF corroborates these findings with trends towards poorer survival for age (>55 years), multifocal cutaneous and extracutaneous LCTMF (including visceral, nodal and blood involvement). Interestingly, LCTMF was not a predictor of OS in the CLIPi data, ⁴³ unlike other published studies, ⁷ , ¹⁰ , ²⁰ , ³⁰ , ³¹ but was an independent predictor for disease progression. ⁴³ While CLIPi serves as a promising starting point for predicting survival, later studies have reported conflicting results. ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ For further validation, the global PROspective Cutaneous Lymphoma International Prognostic Index (PROCLIPI) is anticipated to advance our understanding of the risks for progression to LCTMF. ³⁰

Despite the more aggressive clinical course observed in LCTMF, there is paucity of published treatment‐related data to inform clinical practice guidelines on the optimal management of LCTMF. Encouragingly, the recent National Comprehensive Cancer Network (NCCN) guidelines for MF/SS outline a separate treatment pathway for patients with LCTMF. ³³ However, other international guidelines for the diagnosis and management of MF/SS, including those from the EORTC, Australasian Lymphoma Alliance (ALA) and UK Cutaneous Lymphoma Group, do not presently specify a distinct management approach for LCTMF. ⁴⁸ , ⁴⁹ , ⁵⁰ Reflecting the lack of consensus, treatment heterogeneity was observed in our study, with most patients receiving multimodality management: 84% of patients received systemic therapy and 72% received radiotherapy during their treatment journey. Given the clinical, treatment and prognostic variability in LCTMF, risk‐adapted therapeutic approaches are required. International collaborations are presently underway to investigate a prognostic index and explore management pathways for LCTMF in parallel with underlying classical MF/SS. ³⁰ Furthermore, we encourage prospective investigation of risk‐based treatment protocols to improve outcomes for patients with LCTMF.

We acknowledge the challenges of retrospective data collection, particularly in distinguishing the impact of LCTMF on treatment decisions, alongside management of underlying MF/SS. Treatment heterogeneity is compounded by lacking international consensus on optimal treatment approach for LCTMF. The impact of first‐line therapies on survival could not be interrogated due to the treatment heterogeneity and small patient groups. Evaluation of prognostic outcomes by different CD30+ expression thresholds could not be undertaken due to limited patient numbers, and insufficient reporting of serum LDH at the time of LCTMF diagnosis prevented further investigation. The strength of this study, however, comes from the stable multidisciplinary team within a dedicated CTCL quaternary service, and the mandatory requirement for clinicopathological diagnosis and central review of histopathological diagnoses.

CONCLUSION

For patients with LCTMF, prognosis is poor, with median OS of 3.5 years. Unique to the literature, three distinct clinical and prognostic groups of patients with LCTMF are suggested. Patients with unifocal cutaneous LCTMF were associated with a more favourable prognosis (median OS: 4.6 years) while those with multifocal cutaneous‐only LCTMF and extracutaneous LCTMF were associated with poorer outcomes (2.5 years and 1.1 years respectively). Treatment patterns were highly variable, likely due to concurrent management of the underlying non‐transformed MF/SS and lack of international consensus bespoke to LCTMF. Most patients received multimodality treatment, including both radiotherapy and systemic therapy during their treatment journey. These data support reconsideration of LCTMF in the next iterations of the staging and classification systems for MF/SS.

AUTHOR CONTRIBUTIONS

Designed the research: BAC and HMP. Performed the research: MH, PJ, SL and BAC. Analysed and interpreted the data: PJ and BAC. Performed statistical analysis: BAC and PJ. Wrote the manuscript: PJ, BAC, HMP, SL, CM, CvdW, FB, OB and PB.

CONFLICT OF INTEREST STATEMENT

PJ, MH and SL: Nil. HMP: Advisory boards and/or honoraria received from Kyowa Kirin, MundiPharma, Mallinkrodt, Innate Pharma. CJM: Advisory board and honoraria from Kyowa Kirin. CvdW: Advisory board and honoraria from Kyowa Kirin. FB: Advisory board from Juniper biologics. OB: Advisory board and/or honoraria from Kyowa Kirin, Juniper biologics. PB: Advisory boards and/or honoraria received from AbbVie, Adaptive Biotechnologies, AstraZeneca and Roche Diagnostics, Speakers bureau participation for AstraZeneca and Janssen. BAC: Advisory board and honoraria from Kyowa Kirin.

ETHICS APPROVAL STATEMENT

Ethical approval was obtained from the Human Research Ethics Committee at Peter MacCallum Cancer Centre on 28 February 2022.

PERMISSION TO REPRODUCE MATERIAL FROM OTHER SOURCES

Table S1: Reprinted from Blood, volume 110, issue 6, page 1713–22 ‘ISCL/EORTC. Revisions staging classification mycosis fungoides Sézary syndrome: proposal International Society Cutaneous Lymphomas (ISCL) Cutaneous Lymphoma Task Force European Organization Research Treatment Cancer (EORTC)’. Copyright 2007, with permission from Elsevier.

Supporting information

Data S1.

Johnstone P, Higgins M, Prince HM, Lade S, McCormack C, van der Weyden C, et al. Large‐cell transformation of mycosis fungoides: Patterns of care and patient outcomes. Br J Haematol. 2025;207(3):824–833. 10.1111/bjh.20225

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.