A, A solitary patch with associated alopecia in left eyebrow; B, grouped follicular papules on the trunk; C, alopecia and ulcerating tumors on the scalp; and D, plaques in the neck showing follicular accentuation and extensive secondary bacterial infection (arrows). FMF indicates folliculotropic mycosis fungoides.
A, Histologic examination of a plaque on the left cheek of the first patient shows mucin depositions (follicular mucinosis) and (B) a sparse perifollicular and intrafollicular infiltrate. C, Detail of the infiltrate shows small neoplastic T cells with pleomorphic nuclei. D, Histologic examination of a plaque on the left temple of the second patient shows dense intrafollicular infiltrates with (E,F) a predominance of medium-sized to large neoplastic T-cells. FMF indicates folliculotropic mycosis fungoides.
A and B, OS and DSS of patients with early and advanced plaque-stage FMF. C and D, OS and DSS of patients with early skin-limited FMF (group A), advanced skin-limited FMF (group B), and FMF presenting with extracutaneous disease (group C). DSS indicates disease-specific survival; FMF, folliculotropic mycosis fungoides; OS, overall survival.
eTable 1. Overall survival, disease specific survival and progression free survival in FMF
eTable 2. Uni- and multivariate analysis of DSS, OS and PFS in 203 patients with FMF
Customize your JAMA Network experience by selecting one or more topics from the list below.
van Santen S, Roach REJ, van Doorn R, et al. Clinical Staging and Prognostic Factors in Folliculotropic Mycosis Fungoides. JAMA Dermatol. 2016;152(9):992–1000. doi:10.1001/jamadermatol.2016.1597
Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
Large case series suggest that patients with folliculotropic mycosis fungoides (FMF) have a worse prognosis than patients with classic mycosis fungoides (MF). However, recent studies described a subgroup of patients with FMF with a more favorable prognosis. Distinction between indolent and aggressive FMF may have important therapeutic consequences but is hampered by the inability of the current tumor-node-metastasis-blood (TNMB) staging system to classify patients with FMF in a clinically meaningful way.
To differentiate between indolent and aggressive FMF using clinicopathological criteria and to define prognostic factors in patients with FMF.
Design, Setting, and Participants
In this prospective cohort study, we followed 203 patients with FMF, included in the Dutch Cutaneous Lymphoma Registry between October 1985 and May 2014 at a tertiary referral center hosting the Dutch Cutaneous Lymphoma Registry. Overall, 220 patients with FMF had been registered, but 17 patients with incomplete follow-up data or a history of classic MF were excluded.
Main Outcomes and Measures
Main outcomes included clinical and histological characteristics, disease progression, and survival. Prognostic factors were investigated using Cox proportional hazard regression analysis. Distinction between early plaque-stage FMF and advanced plaque-stage FMF was made by a blinded review of skin biopsy specimens from patients presenting with plaques.
In a cohort of 147 men and 56 women (median [range] age, 59 [15-93] years), patients with histologically early plaque-stage FMF had a very similar overall survival (OS) rate to patients with only patches and/or follicular papules (10-year OS, 71% vs 80%), while the survival rate of patients with histologically advanced plaque-stage FMF was almost identical to that of patients presenting with tumors (10-year OS, 25% vs 27%). Subsequently, 3 clinical subgroups with significantly different survival data were distinguished: early skin-limited FMF (group A; n = 84; 5-year and 10-year OS, 92% and 72%); advanced skin-limited FMF (group B; n = 102; 5-year and 10-year OS, 55% and 28%); and FMF presenting with extracutaneous disease (group C; n = 17; 5-year and 10-year OS, 23% and 2%). Age at diagnosis, large cell transformation and secondary bacterial infection were independent risk factors for disease progression and/or poor survival.
Conclusions and Relevance
The results of this study provide useful criteria to differentiate between indolent and aggressive FMF and confirm the existence of a subgroup of FMF with a favorable prognosis.
In the last decade, folliculotropic mycosis fungoides (FMF) has been widely recognized as a distinct variant of mycosis fungoides (MF).1-9 Clinical and histologic features characteristic of FMF but not or uncommonly found in classic MF include: (1) the histological presence of folliculotropic instead of epidermotropic neoplastic infiltrates, with or without follicular mucinosis1,3-13; (2) preferential localization of skin lesions in the head and neck region, with eyebrow involvement and concurrent alopecia as the most characteristic feature1,3-7,9-11,14,15; (3) the presence of grouped follicular papules, alopecia, acneiform, and cystic lesions1-7,9,11,13; and (4) the frequent occurrence of pruritus and secondary bacterial infections.2,4-7,9 In addition, patients with FMF were less responsive to several first-line skin-directed therapies used in classic MF, such as psoralen plus ultraviolet light A, and had a worse prognosis as compared to classic MF.1,3,5-9 In our initial study, 51 patients with FMF were compared with 158 patients with classic MF, including 122 patients with plaque-stage disease (T2N0M0; stage IB) and 36 patients with tumor-stage MF (T3N0M0). Survival data of patients with FMF were significantly worse than patients with plaque-stage classic MF and similar to patients with tumor-stage MF, although only 14 of 51 patients had tumors or nodules at the time of diagnosis.9 These observations were confirmed in subsequent studies and clearly indicate that the clinical staging system (tumor-node-metastasis-blood [TNMB] staging system) used for MF cannot be used to classify FMF, in particular for patients with plaques, in a clinically meaningful way.3,4,16 The worse prognosis of FMF is also supported by large retrospective cohort studies17-19 of patients with MF, showing that FMF is an independent risk factor for disease progression and lower survival. Because of its characteristic clinicopathologic features and worse prognosis, FMF was included as a distinct entity in recent cutaneous lymphoma classifications.20,21
However, recent studies focused attention on a subgroup of patients with FMF with a favorable prognosis.5,10,11 Distinction between patients with FMF with an indolent and an aggressive clinical disease course may have important therapeutic consequences, but criteria for this indolent subgroup are still ill-defined. In the present study, we reviewed the clinical and follow-up data of 203 patients with FMF who had been followed prospectively after inclusion in the Dutch Cutaneous Lymphoma Registry. The main goal of this study was to develop criteria which could be used to differentiate between indolent and aggressive FMF. In addition, the large size of our cohort offered the opportunity to define additional risk factors for poor patient outcomes. Detailed treatment results and treatment recommendations will be reported separately.
Question How can patients with an indolent form of folliculotropic mycosis fungicides (FMF) be differentiated from those with an aggressive form of the disease?
Findings Clinicopathologic correlation in a cohort of 203 patients with FMF resulted in the recognition of 3 disease subgroups with significantly different survival: patients presenting with patches and/or follicular papules or plaques with histologically sparse perifollicular infiltrates demonstrated highest overall 10-year survival, followed by patients presenting with plaques with histologically dense perifollicular infiltrates, tumors, or erythroderma. Patients with extracutaneous FMF demonstrated the worst overall 10-year survival.
Meaning Clinical presentation and density of perifollicular infiltrate help to differentiate indolent FMF from aggressive FMF.
Between October 1985 and May 2014, 220 patients with FMF were included in the registry of the Dutch Cutaneous Lymphoma Group; 17 of 220 were excluded: 8 patients had incomplete follow-up data, 6 patients had histories of classic epidermotropic MF for 2 to 9 years before they developed skin tumors on the face or scalp with the histologic features of FMF, and 3 patients presented with a solitary skin lesion, in whom another type of cutaneous T-cell lymphoma (in particular a primary cutaneous anaplastic large cell lymphoma) could not be excluded. The final study group consisted of 203 patients. Forty-one of these 203 patients had also been included in our initial study.9 The study was performed in accordance with the Declaration of Helsinki and the Dutch Code for Proper Secondary Use of Human Tissue, approved by the ethics committee of the Leiden University Medical Center.
In all cases, the diagnosis was made by an expert panel of dermatologists and pathologists at one of the regular meetings of the Dutch Cutaneous Lymphoma Group, and all cases met the diagnostic criteria of FMF as described previously.9,20,21 The date of the first diagnostic biopsy was considered the time of diagnosis.
Clinical records, clinical illustrations (available in more than 85% of patients), and follow-up data that had been collected yearly were evaluated. The following variables were recorded: age; sex; duration of skin lesions before diagnosis; type and preferential localization and extent of skin lesions; stage of disease according to the revised TNMB system16; presence of pruritus and secondary bacterial infection; presence of follicular mucinosis and presence of large cell transformation either at diagnosis or during follow-up according to established criteria18; type and result of initial therapy; date and type of disease progression; and date of last contact or death (if applicable).
Disease progression was defined by the development of clinically overt tumors in patients previously having only patch- or plaque stage disease, the development of histologically documented nodal involvement in patients with previously only skin-limited disease, the development of visceral involvement in patients with previous skin and/or nodal disease, and death due to lymphoma. Histological lymph node involvement was assessed using the ISCL-EORTC classification system.16
While patients with only patches, follicular papules, acneiform lesions, or keratosis pilaris-like lesions clearly have early-stage disease (stage IA-IIA) and patients presenting with nodules, tumors, erythroderma and/or extracutaneous disease have advanced FMF (stage IIB-IV), it is uncertain if patients presenting with plaques should be classified as early-stage FMF (stage IA-IB) or as advanced-stage FMF (stage IIB).9 For the purpose of our study, skin biopsy specimens from all patients presenting with plaques were reviewed in a blinded fashion. Lesions histologically showing sparse perifollicular and/or intrafollicular infiltrates containing relatively few and mainly small neoplastic T cells were considered early plaque-stage FMF, while plaques showing more extensive confluent or diffuse infiltrates containing many often medium-sized tumor cells to large tumor cells were considered to have advanced plaque-stage FMF.
The following parameters were analyzed for their prognostic significance in FMF: sex, age at diagnosis (≤60 vs >60 years), duration of skin lesions prior to diagnosis (<12 months vs 12-60 months vs >60 months), extent of skin lesions (solitary vs regional vs generalized); large cell transformation in the first diagnostic biopsy, presence of follicular mucinosis, pruritus, secondary bacterial infection (no vs focal vs extensive), and clinical stage. The term extensive secondary bacterial infection was used for patients with widespread honey-colored crusted lesions at first presentation (Figure 1D).
All statistical analyses were performed using the SPSS statistical software (IBM Corp). Disease-specific survival (DSS) was calculated from the date of first diagnostic biopsy specimen examination until death as result of lymphoma or date of last follow-up. Overall survival (OS) was calculated from the date of diagnosis until patient death from any cause or date of last follow-up. Progression-free survival (PFS) was calculated from the date of diagnosis to the time of disease progression or date of last follow-up. Survival curves were estimated by the method of Kaplan and Meier and comparison between curves was done by log-rank testing. Univariate analysis of parameters with possible prognostic significance for DSS, OS, or PFS was performed using Cox proportional hazards regression analysis. Factors significant at the .05 level in univariate analysis and age at diagnosis, regardless of statistical significance, were included in a multivariate analysis model. In this model P values below .05 were considered significant. To compare clinical outcome parameters among different subgroups of patients with FMF a χ2 test for goodness of fit was performed for categorical variables and a Kruskal-Wallis test was performed to compare medians between subgroups; P values below .05 were considered significant for both tests.
The main clinical characteristics are summarized in Table 1. The study included 147 men and 56 women (male to female ratio: 2.6). The median (range) age at diagnosis was 59 (15-93) years. Only 3 of 203 patients (1.5%) were 18 years or younger. Overall, 174 patients (86%) presented with generalized skin lesions involving multiple body regions, while 9 (4%) and 20 (10%) patients had presented with solitary or localized skin lesions, respectively. Thirteen patients showed erythroderma at first presentation. This group included 6 patients with stage III (skin-limited disease) and 7 patients with stage IV disease, 4 of whom met the criteria for peripheral blood involvement of Sézary syndrome (SS).
In 129 of 203 patients (64%), skin lesions were preferentially located and most pronounced in the head and neck area (Figure 1). Infiltrated plaques in the eyebrows with concurrent alopecia were observed in more than 50% of patients (n = 110 of 193 patients [10 cases unknown]), either at the time of diagnosis or during follow-up. Three patients presented with a leonine face. Only 20 of 203 (10%) patients had no skin lesions in the head and neck area at first presentation. Associated alopecia on affected skin sites was seen in about 80% of patients (n = 150 of 186 patients [17 cases unknown]). Secondary bacterial infection was observed in 43 patients (21%) and was extensive in 21 patients. Almost 80% of patients complained of moderate to severe pruritus (n = 133 of 173 patients [30 cases unknown]).
Four patients had a history of Hodgkin lymphoma, 9 to 22 years before the diagnosis FMF was made. Three patients had a coexistent hematological disorder: 1 patient had B-cell chronic lymphocytic leukemia, 1 patient had myelodysplastic syndrome, and 1 patient had essential thrombocytosis.
At time of diagnosis 186 patients (92%) had skin-limited disease, while 17 patients (8%) had nodal or visceral disease at first presentation (stage IV). Among those patients with skin-limited disease, 67 patients presented with only patches, follicular papules, acneiform or keratosis pilaris-like lesions (stage IA-IIA); 55 patients presented with nodules or tumors (stage IIB); 6 patients presented with erythroderma (stage III); and 58 patients presented with plaques. As noted before, it is uncertain on the basis of clinical evaluation alone if patients presenting with plaques should be classified as stage IA to IIA or stage IIB. Based on histologic criteria distinction was therefore made between patients presenting with early plaque-stage disease (n = 17) and patients with advanced plaque-stage disease (n = 41) (Figure 2). The OS and DSS of the whole group of patients presenting with plaques (n = 58) were intermediate between those of patients presenting with only follicular papules and/or patches and those of patients presenting with nodules, tumors, or erythroderma (eTable 1 in the Supplement). However, patients presenting with histologically classified early plaque-stage FMF had almost identical survival and progression data as patients presenting with only patches and follicular papules, while patients presenting with histologically classified advanced plaque-stage FMF had a very similar course to patients presenting with tumors or nodules. The differences in 10-year OS (80% vs 25%; P = .004) and 10-year DSS (100% vs 35%; P = .006) between patients with early plaque-stage and advanced plaque-stage FMF were highly significant, which confirms the usefulness of the clinicopathologic approach used (Figure 3A and B). Taken together, 3 clinically relevant stages were distinguished: patients with early skin-limited disease (n = 84; group A), patients with advanced skin-limited disease (n = 102; group B), and patients presenting with extracutaneous disease (n = 17; group C) (Figure 3C and D).
Both in univariate and multivariate analysis, age at diagnosis older than 60 years, clinical stage (as defined above), and the presence of extensive secondary bacterial infection were independent factors associated with reduced OS, DSS, and PFS, while sex, the extent of skin lesions, the duration of skin disease prior to diagnosis, follicular mucinosis, and the presence of pruritus had no effect on survival or disease progression (eTable 2 in the Supplement). The presence of large cell transformation at first presentation was independently associated with reduced PFS but not with reduced OS. The relation between large cell transformation and DSS was not significant (P = .05).
Large cell transformation at first presentation was observed in 33 patients, including 21 cases in group B, 10 cases in group C, and only 2 in group A. The median survival of these 33 cases was 32 months, the 5- and 10-year OS was 40% and 20%, respectively, and the 5- and 10-year DSS was 41% and 20%, respectively.
Extensive secondary bacterial infection was particularly found in patients with advanced FMF (19 of 21 patients), and in most cases affected skin lesions were in the head and neck region (17 of 19 patients) (Figure 1D). The median survival of these 21 patients was only 22 months, and the 5-year OS and DSS were 20% and 26%, respectively.
After initial therapy, 25 of 203 (12%) patients never had a relapse, and the median (range) duration of this sustained complete remission was 68 (12-169) months; 101 patients (50%) showed continued disease without progression to a higher stage, while 77 patients (38%) showed disease progression (Table 2). Altogether, 25 patients presented with visceral involvement (n = 1) or developed visceral involvement. Visceral sites most commonly affected were the central nervous system (n = 11), lungs (n = 9), bone marrow (n = 6), and oral or nasal mucosa (n = 5), while 3 patients developed peripheral blood involvement.
After a median (range) follow-up of 51 (3-260) months, 32 of 203 patients were in complete remission, 78 patients were alive with disease, 59 patients died of lymphoma, and 34 patients died of unrelated disease. For those 203 patients the 5- and 10-year OS were 67% and 45%, respectively, and the 5- and 10-year DSS were 75% and 60%, respectively, with significant differences between the 3 clinical subgroups (Table 2) (Figure 3C and D).
Among 84 patients with early skin-limited disease (group A), 17 patients (20%) developed skin tumors and 2 patients (2%) developed extracutaneous disease. Five (6%) patients died of lymphoma after a median (range) follow-up of 76 (32-198) months. The clinical presentation and histology of these 5 cases did not differ from other patients in group A. Interestingly, all 5 patients presented with widespread follicular papules and patches, and none of them presented with plaques.
While previous studies emphasized the worse prognosis of FMF as compared with classic type MF, more recent studies focused attention on a subgroup of FMF with a favorable prognosis.5,10,11 However, criteria to differentiate between indolent and more aggressive FMF are still ill-defined, which is mainly caused by the inability of the TNMB system to classify FMF patients presenting with plaques in a clinically meaningful way.3,4,16 The results of the present study, in which additional histologic criteria were used to distinguish early plaque-stage FMF from advanced plaque-stage FMF, provide useful criteria to differentiate between indolent and aggressive FMF. Multivariate analysis showed that clinical stage, age at diagnosis, large cell transformation, and extensive secondary bacterial infection were independent risk factors for disease progression and/or poor survival.
In classic MF, clinical stage according to the revised TNMB criteria is the most important factor in predicting survival and risk of disease progression.19,22 In our study, patients with FMF presenting with only patches and/or follicular papules (stage IA-IIA) had indeed an excellent prognosis with a 5-year OS and DSS of 92% and 95%, respectively, while patients presenting with tumors and/or nodules (stage IIB) had a 5-year OS and DSS of 50% and 59%, respectively (eTable 1 in the Supplement). The favorable prognosis of patients presenting with only patches and/or follicular papules is consistent with recent literature.5,10,11 It is however unclear if FMF patients presenting with plaques should be classified as stage IA, IIA, or stage IIB.9 Clinically identical plaques may be caused by dense neoplastic infiltrates but also by excessive mucin depositions or an extensive inflammatory infiltrate. In the present study, distinction was therefore made between plaques histologically characterized by sparse intrafollicular and/or perifollicular infiltrates containing relatively few and mainly small neoplastic T cells (early plaque-stage FMF) and plaques histologically showing more extensive confluent or diffuse infiltrates containing many often medium-sized to large tumor cells (advanced plaque-stage FMF). Patients with histologically classified as early plaque-stage FMF had an almost equal clinical course and survival to patients presenting with only patches and/or follicular papules, while patients with histologically classified advanced plaque-stage FMF had a very similar course to patients presenting with tumors, confirming the usefulness of this clinicopathologic approach (eTable 1 in the Supplement). Our results validated those of Hodak et al presented at the 2014 meeting of the EORTC Cutaneous Lymphoma Group in Paris. Using a very similar clinicopathologic approach and very similar histologic criteria, distinction could be made between early-stage and advanced-stage FMF (E. Hodak; personal email communication; February 11, 2016). Whether patients with advanced plaque-stage disease should be classified as tumor-stage disease (T3 score; stage IIB) rather than plaque-stage disease (T1B/T2B score; stage IA-IIA), as suggested by Hodak et al, is a matter of debate. The similar survival and progression rates between patients with advanced plaque-stage FMF and tumor-stage FMF seem to justify such an upgrading. However, in the revised TNMB system, skin score is determined by clinical presentation and skin lesions histologically classified as either early or advanced plaque-stage FMF may be clinically indistinguishable (Figure 2).16 Additional studies investigating the reproducibility of histology-based distinctions between early and advanced plaque-stage FMF should be awaited before further revisions of the TNMB system are made. A meeting of the EORTC Cutaneous Lymphoma Pathology Group, in which the reproducibility of a histologic stratification is investigated by an independent panel of (dermato)pathologists is scheduled for 2017.
Previous studies6,7 have been unsuccessful to detect independent prognostic parameters, probably because of the relatively small number of patients included. In the present study, in addition to clinical stage, age older than 60 years, large cell transformation, and the presence of extensive secondary bacterial infection at the time of first presentation were independent factors associated with reduced OS, DSS, and/or PFS. Clinical stage of disease, age, and large cell transformation are well-known risk factors in classic MF.19,22-24 The presence of secondary bacterial infection as an adverse risk factor has not been found before. It is well known that bacterial toxins acting as superantigens may stimulate the proliferation of malignant T cells and may worsen disease in patients with a cutaneous T-cell lymphoma.25,26 However, whether this or other mechanisms are responsible for the poor outcome in patients with FMF with extensive secondary bacterial infection is unknown.
The results of the present study confirm the existence of a substantial subgroup of patients with FMF with a good prognosis. Apart from patients presenting with only patches and/or follicular papules, this group also contains patients with plaques histologically characterized by sparse intrafollicular and/or perifollicular infiltrates containing relatively few and mainly small neoplastic T cells. Distinction between patients with indolent and aggressive FMF is important because it may have therapeutic consequences.
Corresponding Author: Suzanne van Santen, MD, Leiden University Medical Center, Department of Dermatology, B1-Q Albinusdreef 2, 2333 ZA Leiden, the Netherlands (email@example.com).
Accepted for Publication: April 18, 2016.
Published Online: June 8, 2016. doi:10.1001/jamadermatol.2016.1597.
Author Contributions: Drs van Santen and Willemze had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: van Santen, Roach, Horváth, de Pooter, Vermeer, Willemze.
Acquisition, analysis, or interpretation of data: van Santen, Roach, van Doorn, Bruijn, Sanders, van Rossum, de Haas, Veraart, Bekkenk, Vermeer, Willemze.
Drafting of the manuscript: van Santen, Roach, de Pooter, Willemze.
Critical revision of the manuscript for important intellectual content: van Santen, Roach, van Doorn, Horváth, Bruijn, Sanders, van Rossum, de Haas, Veraart, Bekkenk, Vermeer, Willemze.
Statistical analysis: van Santen, Roach, van Doorn, Willemze.
Administrative, technical, or material support: van Santen, Roach, Horváth, Sanders, de Pooter, van Rossum, de Haas, Veraart, Vermeer, Willemze.
Study supervision: Roach, van Doorn, Sanders, Bekkenk, Vermeer, Willemze.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by the Department of Dermatology, Leiden University Medical Center, Leiden, The Netherlands.
Role of the Funder/Sponsor: The funder/sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.