Association of a Proposed New Staging System for Folliculotropic Mycosis Fungoides With Prognostic Variables in a US Cohort | Dermatology | JAMA Dermatology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Clinical Morphologic Features in Folliculotropic Mycosis Fungoides
Clinical Morphologic Features in Folliculotropic Mycosis Fungoides

A, Perifollicular papules coalescing into plaques on the flank in a patient with early-stage cutaneous disease. B, Comedone-like papules and cystic nodules on the face of a patient with early-stage cutaneous disease. C, Multiple plaques without follicular prominence on buttocks in a patient with advanced-stage cutaneous disease.

Figure 2.  Histopathologic Features Seen in Folliculotropic Mycosis Fungoides (FMF)
Histopathologic Features Seen in Folliculotropic Mycosis Fungoides (FMF)

A, Low-power image of early-stage cutaneous FMF with perifollicular and intrafollicular sparse to moderate lymphocytic infiltrate (hematoxylin-eosin, original magnification ×40). B, Medium-power image of early-stage cutaneous FMF demonstrating mucin within follicular epithelium (hematoxylin-eosin, original magnification ×200). C, High-power image of early-stage cutaneous FMF demonstrating penetration of lymphocytes into follicular epithelium (hematoxylin-eosin, original magnification ×400). D, Low-power image of advanced-stage cutaneous FMF with a dense and deep perifollicular and intrafollicular lymphocytic infiltrate (hematoxylin-eosin, original magnification ×40). E, Medium-power image of advanced-stage cutaneous FMF demonstrating follicular prominence of the infiltrate despite the diffuse configuration (hematoxylin-eosin, original magnification ×200). F, High-power image of advanced-stage cutaneous FMF demonstrating penetration of lesional lymphocytes into the follicular epithelium and follicular distortion (hematoxylin-eosin, original magnification ×400).

Figure 3.  Kaplan-Meier Plots for Estimated Overall and Disease-Specific Survival (DSS)
Kaplan-Meier Plots for Estimated Overall and Disease-Specific Survival (DSS)

A, For early-stage cutaneous disease, estimated 5-year overall survival rate was 96% (95% CI, 89%-103%); 10-year rate, 82% (95% CI, 65%-98%); and 15-year rate, 65% (95% CI, 33%-97%). For advanced-stage cutaneous disease, estimated 5- and 10-year overall survival rates were 70% (95% CI, 41%-98%); 15-year rate, 53% (95% CI, 16%-89%). B, For early-stage cutaneous disease, estimated 5-, 10-, and 15-year disease-specific survival rates were 96% (95% CI, 89%-103%). For advanced-stage cutaneous disease, estimated 5- and 10-year DSS rates were 70% (95% CI, 41%-98%); 15-year rate, 53% (95% CI, 16%-89%).

Table 1.  Clinical, Histopathologic, and Molecular Features of Patients With FMF
Clinical, Histopathologic, and Molecular Features of Patients With FMF
Table 2.  Cox Proportional Hazards Regression Model Data
Cox Proportional Hazards Regression Model Data
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    December 9, 2020

    Association of a Proposed New Staging System for Folliculotropic Mycosis Fungoides With Prognostic Variables in a US Cohort

    Author Affiliations
    • 1Cutaneous Hematopathology Clinic, Department of Dermatology, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City, Mexico
    • 2Department of Dermatology, California Pacific Medical Center, San Francisco
    • 3Dermatopathology Service, Departments of Pathology and Dermatology, University of California, San Francisco
    • 4University of California, San Francisco Helen A. Diller Family Comprehensive Cancer Center
    • 5Department of Epidemiology and Biostatistics, University of California, San Francisco
    • 6Division of Hematology and Oncology, Department of Medicine, University of California, San Francisco
    JAMA Dermatol. 2021;157(2):157-165. doi:10.1001/jamadermatol.2020.4372
    Key Points

    Question  Does a new cutaneous staging system for folliculotropic mycosis fungoides (FMF) better estimate survival compared with the staging system for conventional disease?

    Findings  In this cohort study of 42 patients with FMF, age was associated with overall survival, and advanced-stage disease was associated with disease-specific survival. Subdividing FMF into early and advanced cutaneous stages effectively was associated with survival in this cohort.

    Meaning  These findings suggest that FMF is a heterogeneous disease with early and advanced cutaneous stages; that the new staging system is effective in estimating survival in a US cohort; and that the poor prognosis initially associated with FMF only applies to the advanced cutaneous stage.

    Abstract

    Importance  A new cutaneous staging system for folliculotropic mycosis fungoides (FMF) has been purported to better estimate survival compared with the staging system for conventional mycosis fungoides.

    Objective  To analyze predictive variables associated with survival and evaluate the effectiveness of the newly proposed staging system for estimating overall survival and disease-specific survival (DSS) in a US cohort.

    Design, Setting, and Participants  This cohort study assessed 195 patients with FMF in the dermatopathology database of the University of California, San Francisco from January 1, 1990, to April 31, 2012, for eligibility. A total of 153 patients were excluded for the following reasons: (1) alternative diagnoses were favored ranging from benign dermatitides to other forms of cutaneous lymphoma; (2) technical problems with slides; and (3) lack of follow-up information. Data were analyzed from January 1, 2018, to August 31, 2020.

    Main Outcomes and Measures  Kaplan-Meier curves were used to estimate overall survival and DSS for the entire cohort. Possible predictive variables associated with survival were evaluated using Cox proportional hazards regression modeling. Each variable was examined separately, followed by a multiple-variable model. Kaplan-Meier curves were used to estimate overall survival and DSS by subdividing the cohort into early- and advanced-stage cutaneous disease.

    Results  Forty-two patients were included in the analysis (mean age at diagnosis, 55 [range, 8-89] years; 31 men [74%]). For the entire cohort, the estimated 5-year overall survival rate was 89% (95% CI, 79%-99%); 10-year rate, 78% (95% CI, 63%-92%); and 15-year rate, 58% (95% CI, 31%-85%). Estimated 5- and 10-year DSS rates were 89% (95% CI, 79%-99%); 15-year rate, 80% (95% CI, 61%-99%). For overall survival in the multiple-variable Cox proportional hazards regression model, only age was statistically significant (hazard ratio [HR] per 10-year age increase, 3.1; 95% CI, 1.4-7.2; P = .008), whereas for DSS, only cutaneous disease was statistically significant (HR, 11.4; 95% CI, 1.3-103.0; P = .03). When stage stratified, the 5-year estimated overall survival rate for early-stage disease was 96% (95% CI, 89%-103%); 10-year rate, 82% (95% CI, 65%-98%); and 15-year rate, 65% (95% CI, 33%-97%). For advanced-stage disease, the estimated 5- and 10-year overall survival rates were 70% (95% CI, 41%-98%); the 15-year rate, 53% (95% CI, 16%-89%). For early-stage cutaneous disease, the estimated 5-, 10- and 15-year DSS rates were all 96% (95% CI, 89%-103%). For advanced-stage cutaneous disease, the estimated 5- and 10-year DSS rates were 70% (95% CI, 41%-98%); the 15-year rate, 53% (95% CI, 16%-89%).

    Conclusions and Relevance  Cox proportional hazards regression modeling demonstrated cutaneous stage to be the only statistically significant predictive variable associated with DSS. Subdividing FMF into early and advanced cutaneous stages was associated with effective estimation of survival in this cohort. Thus, findings suggest that FMF is a heterogeneous disease with early and advanced cutaneous stages; that the new staging system is effective in estimating survival in a US cohort; and that the poor prognosis initially associated with FMF only applies to the advanced cutaneous stage.

    Introduction

    Folliculotropic mycosis fungoides (FMF) was recognized as a distinct entity and distinguished from conventional mycosis fungoides (CMF) by its clinicopathologic features and purportedly worse prognosis.1-8 Differing survival data for FMF have been reported. In 2002, the first retrospective cohort study of patients with FMF4 reported a poor prognosis when compared with stage-matched patch- or plaque-stage CMF, with 5- and 10-year disease-specific survival (DSS) rates of 68% and 26%, respectively, and 5- and 10-year overall survival rates of 64% and 14%, respectively. In 2008, a US study5 also reported poor survival, with 10- and 15-year overall survival rates of 82% and 41%, respectively. However, later non-US studies described better survival,9-14 including a few reporting overall survival as favorable as 94%11 to 100%.10,12 It has been hypothesized that such discrepancies partially stem from using different FMF definitions and therefore that studies have analyzed different cohorts of patients. In particular, it has been suggested that studies reporting a poorer prognosis included mostly patients with advanced-stage disease, whereas others demonstrating a good prognosis included mostly patients with early-stage disease. Therefore, a new FMF-specific staging system, which divides patients between early and advanced cutaneous stages, has been proposed.15 This system, currently lacking full validation, is considered potentially better associated with survival in FMF compared with the tumor-node-metastasis-blood (TNMB) system for CMF. Thus, we sought to test its utility in estimating survival in our institutional cohort of patients with FMF. In addition, although the implicit assumption behind using cutaneous stage as the parameter is that stage is best associated with survival, we wanted to test this assumption in our cohort by performing Cox proportional hazards regression modeling to examine multiple variables associated with overall survival and DSS in the entire cohort.

    Methods
    Patients

    The Committee on Human Research from the University of California, San Francisco (UCSF) granted initial approval for this study. The institutional review board of the UCSF approved the present study and waived the need for informed consent for this retrospective evaluation of medical records. This cohort study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

    The definition of FMF incorporated clinical and histopathologic features. Clinical morphologic features included erythematous plaques with or without follicular prominence, comedone-like papules or pustules, and/or cystlike nodules resembling infundibular follicular cysts (acneiform lesions); perifollicular erythematous and scaly papules that did not coalesce into plaques; noncomedonal nodules; and erythroderma. The histopathologic definition consisted of a mostly perifollicular infiltrate of lymphocytes with some penetrating the follicular epithelium, regardless of the presence or absence of mucin. A small amount of epidermotropism was allowed; however, in all cases, folliculotropism dominated.

    For our cohort, we searched the UCSF dermatopathology database (January 1, 1990, to April 31, 2012) for patients with specimens coded as mycosis fungoides, folliculocentric, folliculotropic, or follicular; mycosis fungoides, follicular mucinosis; and follicular mucinosis. We excluded 153 of 195 cases for the following: (1) 1 to 2 lesions on or near the face with sparse infiltrates that did not penetrate mucin-laden follicles (idiopathic follicular mucinosis [IFM]) (n = 7); (2) solely facial lesions with sparse infiltrates, lacking mucin and intrafollicular lymphocytes (rosacea) (n = 9); (3) favoring an alternative diagnosis other than IFM or rosacea (lupus erythematosus or eczema with incidental follicular mucinosis) (n = 10); (4) no retrievable slides or survival information (n = 41); (5) pandermal infiltrates incidentally involving follicles (tumor-stage CMF with incidental follicular involvement) (n = 15); (6) epidermotropism dominated over folliculotropism (CMF with focal follicular involvement) (n = 37); (7) truncal CMF that preceded facial lesions demonstrating follicular involvement (CMF with incidental follicular involvement upon progression) (n = 17); and (8) poor slide quality or slides that were transected such that the reticular dermis could not be evaluated (n = 17). We included cases with concurrent CMF lesions if CMF developed after FMF and if the predominant clinical and histopathologic presentation was that of FMF (n = 6), similar to other studies.4,13

    Clinical information was obtained through medical record review of patients treated at UCSF. For those treated elsewhere but whose slides were interpreted at UCSF, we contacted their dermatologists and/or oncologists and reviewed available medical records. For all patients, we used a standardized questionnaire to obtain the clinical information, and clinical photographs were reviewed when available. Clinical parameters evaluated included age at diagnosis, sex, FMF distribution/morphology, CMF presence, timing of CMF development compared with FMF, and status at last follow-up. Histopathologic parameters included epidermal, dermal, and follicular findings; vascular involvement; density of the infiltrate (sparse, moderate, or marked); and depth (superficial to middermal vs deep if it extended into the deep reticular dermis and/or subcutis). We assessed T-cell clonality using primers to the γ-chain genes of the T-cell receptor. Immunoperoxidase staining and/or in situ hybridization for Epstein-Barr virus were performed on cases showing angioinvasion or angiodestruction, and a colloidal iron stain to assess for follicular mucin deposition was applied in selected cases. Sections were reviewed or verified by 3 dermatopathologists (Y.C.-J., P.E.L., and L.B.P.), and scoring discrepancies were resolved through additional review and consensus judgment.

    For clinicopathologic cutaneous staging of FMF, we classified cases into early vs advanced as follows: early lesions were erythematous plaques with or without follicular prominence, comedone-like papules or pustules and/or cystlike nodules (acneiform lesions) or discrete perifollicular erythematous and scaly papules that did not coalesce into plaques; with sparse to moderate infiltrates of superficial to middermal depth surrounding normal or distorted follicles, comedones or cysts, or dilated infundibula. Advanced lesions were erythematous plaques with or without follicular prominence and/or noncomedonal nodules showing markedly dense and deep folliculocentric infiltrates, angiocentricity or angiodestruction, and/or ulceration. We subclassified plaques into early vs advanced using microscopic features and verified in all cases that biopsy specimens were obtained from the most representative lesions.

    Statistical Analysis

    Data were analyzed from January 1, 2018, to August 31, 2020. Kaplan-Meier curves were used to estimate overall survival and DSS for the entire cohort (considering non-FMF deaths as censored observations). After examining the Kaplan-Meier curves for the complete population, we examined whether any variables were associated with survival. To this end, we fitted a set of single-variable models using Cox proportional hazards to assess whether each individual variable was associated with DSS and overall survival for the entire cohort, including age, sex, ethnicity, cutaneous stage (early vs advanced), and T-cell gene clonality in skin biopsy specimens. Initially, we examined each variable separately in a univariate analysis and subsequently built a multiple-variable model using forward stepwise regression, adding in additional variables one at a time. Given the small numbers, we started with the most statistically significant variable in the univariate analysis and subsequently added each variable to determine whether any were statistically significant when combined with the most statistically significant. We checked proportionality and linearity for all Cox proportional hazards regression models and evaluated for outliers. Estimated hazard ratios (HRs) from Cox proportional hazards regression modeling were given along with 95% CIs and P values to describe the observed effects, associated uncertainties, and strength, where 2-sided P < .05 indicated statistical significance. Next, Kaplan-Meier curves were used to estimate overall survival and DSS stratified by cutaneous stage.

    Results
    Clinical Features

    A total of 42 patients were included in the analysis (31 men [74%] and 11 women [26%]; mean age at diagnosis, 55 [range, 8-89] years). Thirty-one patients had early-stage cutaneous FMF, and 11 had late-stage cutaneous FMF. The predominant morphologic findings in the early-stage group consisted of erythematous plaques with follicular prominence (19 [61%]), whereas in the advanced-stage group, it consisted of erythematous plaques without follicular prominence (6 [55%]). Nodules without follicular prominence were only seen in the advanced-stage group (4 [36%]). Many patients had more than 1 clinical morphologic feature (Table 1 and Figure 1).

    Histopathologic Features

    Follicular mycosis fungoides was characterized by a sparse to moderate perifollicular and intrafollicular lymphoid infiltrate in 31 cases (74%) and markedly dense infiltrate in 11 (26%). The depth of the infiltrate was superficial to middermal in 26 cases (62%) and deep in 11 (26%). It was not possible to evaluate the infiltrate’s depth in 5 biopsy specimens (12%) owing to specimen transection, although for inclusion in the study, biopsy specimens extended at least into the superficial reticular dermis (Table 1 and Figure 2).

    Cutaneous Stage Determination

    Based on our proposed clinicopathologic staging, 11 cases (26%) were advanced-stage cutaneous lesions. The remaining 31 (74%) fulfilled criteria for early-stage cutaneous lesions.

    Survival and Cox Proportional Hazards Regression Modeling

    At last follow-up, 33 patients (79%) were alive, including 19 (45%) with disease, 12 (29%) without disease, and 2 (5%) with uncertain disease status. Nine patients died during follow-up (21%), including 5 of FMF and 4 of the following unrelated events: myocardial infarction (aged 81 years), metastatic lung carcinoma (aged 83 years), complications of a below-the-knee amputation secondary to diabetes (aged 60 years), and a presumed stroke during sleep (aged 82 years). In all who died of unrelated disease, FMF was either stable or in remission. Of the 5 who died of FMF, 4 had advanced-stage cutaneous disease and 1 had early-stage disease.

    For the entire cohort, the estimated 5-year overall survival rate was 89% (95% CI, 79%-99%); 10-year rate, 78% (95% CI, 63%-92%); and 15-year rate, 58% (95% CI, 31%-85%). The estimated 5- and 10-year DSS rates were both 89% (95% CI, 79%-99%); the estimated 15-year DSS rate was 80% (95% CI, 61%-99%).

    Cox proportional hazards regression modeling demonstrated that, in the univariate analysis for the entire cohort, age was a statistically significant variable associated with overall survival (HR per 10-year increase in age, 2.7; 95% CI, 1.3-5.6; P = .007) (Table 2). The multiple-variable model resulted in age alone having a statistically significant association with overall survival (HR for every 10-year increase in age, 3.1; 95% CI, 1.4-7.2; P = .008). Cutaneous stage rendered an HR of 3.7 (95% CI, 0.9-15.5; P = .08). Hence, cutaneous stage was not statistically significant. Therefore, the model with age alone was our final multiple-variable model, with no other variables providing statistically significant added value.

    Regarding DSS, Cox proportional hazards regression modeling demonstrated that, in the univariate analysis for the entire cohort, cutaneous stage was statistically significant, with an HR of 11.4 (95% CI, 1.3-103.0; P = .03). No other variables reached significance. The DSS multiple-variable model yielded the same results as the univariate analysis. Therefore, the model with cutaneous stage alone as a variable was our final multiple-variable model for DSS. Of note, in the multiple-variable model, when age was accounted for, cutaneous stage remained significant with an HR of 12.5 (95% CI, 1.3-119.0; P = .03).

    We additionally separated the cohort into early- and advanced-stage cutaneous disease and analyzed these subgroups separately for overall survival and DSS. For early-stage cutaneous disease, the 5-year overall survival rate was 96% (95% CI, 89%-103%); 10-year rate, 82% (95% CI, 65%-98%); and 15-year rate, 65% (95% CI, 33%-97%). For advanced-stage cutaneous disease, the estimated 5- and 10-year overall survival rates were 70% (95% CI, 41%-98%); the 15-year rate, 53% (95% CI, 16%-89%). The estimated 5-, 10-, and 15-year DSS rates for early-stage cutaneous disease were all 96% (95% CI, 89%-103%). The 5- and 10-year DSS rates for advanced-stage cutaneous disease were 70% (95% CI, 41%-98%); the 15-year rate, 53% (95% CI, 16%-89%) (Figure 3).

    On checking for proportionality and linearity for all Cox proportional hazards regression models, we did not find any statistically significant departures from proportionality. In addition, diagnostic plots did not show clear departures from proportionality or linearity. When examining outliers via multiple diagnostic plots and by fitting models with and without the most extreme outlier(s), we did not note any influential outliers. It should be noted, however, that such influential points are difficult to disentangle for a data set of this size.

    Discussion

    Folliculotropic mycosis fungoides was initially reported to exhibit a prognosis similar to that of tumor-stage CMF.4-8 This finding contrasts with European and Middle Eastern studies published since 2013.10-13 In our series, the archetypal clinical and microscopic findings described for FMF were consistently reproduced except in 1 patient who displayed noticeable angiocentricity and angiodestruction, a finding previously unreported in FMF; we considered this finding a sign of advanced-stage disease. To exclude other angiocentric lymphomas in this case, we performed immunohistochemistry with CD56 and GM3 and in situ hybridization for Epstein-Barr virus, the findings of which were all negative.

    Herein, we initially analyzed overall survival and DSS for the entire cohort. Our Kaplan-Meier analyses revealed a relatively high survival for the first 10 years, with 5- and 10-year overall survival rates of 89% and 78%, respectively. However, at 15 years, the overall survival declined to 58%. The 5- and 10-year DSS rates were high at 89%, and the 15-year DSS rate was high as well at 80%. Therefore, the 5- and 10-year overall survival rates and the 5-, 10-, and 15-year DSS rates are favorable and roughly comparable to those of patients with early-stage CMF16,17 and in keeping with results derived from non-US FMF cohorts published after 2013.10-13 As such, the 58% 15-year overall survival rate can be attributed to increased age because older people are more likely to die than younger people. Therefore, our data support a more favorable survival in FMF compared with initial reports. To our knowledge, this outcome represents the first report of a more favorable survival in patients with FMF in a US cohort.

    When we stratified our cases by cutaneous stage and evaluated survival, we found meaningful differences. In particular, for early-stage cutaneous disease, the estimated 5- and 10-year overall survival rates were 96% and 82%, respectively, and the DSS rates were 96% at both cutoffs. The 15-year overall survival rate for early-stage cutaneous disease was 65%, but the DSS rate was 96%. Thus, the lower 15-year overall survival rate for early-stage cutaneous disease can be accounted for by older people being more likely to die, similar to the low 15-year overall survival for the entire cohort. As such, early-stage cutaneous disease is associated with high DSS. By contrast, for advanced-stage cutaneous disease, the estimated overall survival and DSS rates for 5 and 10 years were all 70%, and those for 15 years were 53%. Therefore, these data suggest that there are significant differences in survival by cutaneous stage, supporting the finding that the poor prognosis initially reported for FMF applies only to advanced-stage cutaneous disease.

    Although the implicit assumption behind using definitions of early and advanced phase as a way to stage FMF is that this substratification is best associated with outcomes, we wanted to test this assumption in our cohort. We did this by performing Cox proportional hazards regression modeling to examine multiple variables associated with overall survival and DSS in the entire cohort. In both univariate and multiple-variable overall survival models, only age reached statistical significance, which likely stems from some deaths being secondary to diseases typically associated with older age. Based on this thinking, the overall survival association with age can be interpreted as older people being more likely to die than younger people. Regarding DSS, only cutaneous stage was statistically significant in both univariate and multiple-variable models. Following an early- vs advanced-stage cutaneous stratification scheme, someone alive with advanced-stage cutaneous disease has an estimated 11.4 times higher probability of dying of FMF than someone with early-stage cutaneous disease. Furthermore, when accounting for age in the DSS multiple-variable model, the association of cutaneous stage with DSS remained statistically significant, indicating that age is not a hidden cause of the lower DSS in patients with advanced-stage cutaneous disease. The statistical significance of cutaneous stage reflects strong evidence of a difference in survival for early- vs advanced-stage cutaneous disease. Therefore, Cox proportional hazards regression modeling data demonstrate that cutaneous stage is the most important variable associated with DSS in our cohort and supports using early vs advanced forms of skin disease as the variable to stage patients.

    The CMF-TNMB classification system based partially on body surface area involved does not seem applicable to FMF because it does not appear to accurately estimate its prognosis. In particular, a patient with FMF might have less than 10% body surface area involved, with thick plaques incorporating a dense and deep infiltrate, and thus would be understaged in the CMF-TNMB staging system.17,18 Of note, noncutaneous parameters included in the CMF-TNMB staging system, such as extracutaneous involvement, may affect the survival of patients with FMF in a similar manner to patients with CMF. Therefore, we advocate that the system for determining clinical stage proposed herein be used to determine only the T portion of the TNMB staging system. Other reports also argued that the CMF-TNMB staging system does not apply to FMF4,5,15 and suggested an alternative system that subdivides patients into early- and advanced-stage cutaneous disease. Hodak et al13 were the first to propose this system, which was subsequently advocated by van Santen et al.15 Of note, both sets of investigators discuss cases that do not easily subcategorize into early- and advanced-stage cutaneous disease. van Santen et al15 identified patients with plaques in which the infiltrate was diffuse, with medium to large lymphocytes, and the prognosis was similar to those with advanced disease. By contrast, those with plaques consisting of sparse infiltrates with small lymphocytes had a prognosis similar to those with early-stage disease. Thus, the authors argue that plaque stage should be determined by histopathology.15 Hodak et al13 also differentiated plaques into stages but through clinical attributes. Because van Santen et al15 demonstrated that, for clinically similar plaques, there could be a different prognosis based on histopathologic findings, we concur with and used their method. Therefore, we recommend performing a deep biopsy that captures the full length of the follicle, such as a 4-mm or larger punch biopsy, if there is a clinical impression of FMF so that the clinical stage can optimally be determined.

    Earlier FMF studies reported a worse survival compared with stage-matched CMF.4-8 A major difference with these previous studies and our cohort is that earlier studies did not subanalyze patients by cutaneous stages. In fact, the European study15 that divided their cases into early- and advanced-stage disease and incorporated some of the same patients included in the first study that reported a poor prognosis4 found a favorable outcome in early-stage cutaneous disease. Alternatively, differences in survival may lie in inclusion criteria. In early studies, cases with a sparse infiltrate but numerous clinical lesions could have been classified as IFM and excluded. The distinction between IFM and FMF is still a point of debate. Some authorities posit that IFM is not an authentic entity and represents a unilesional form of FMF.19 However, in our opinion, IFM is an authentic entity but can only be used as a designation for patients with 1 to 2 lesions on or close to the face and biopsy specimens that demonstrate follicular mucinosis and mild lymphocytic infiltrates. We excluded patients we categorized as having IFM; however, if multiple lesions were present, we classified them as having FMF.

    Limitations

    Limitations to our study include biases associated with its retrospective nature and the relatively small sample. Nonetheless, given the rareness of FMF, single-center studies with larger numbers are challenging. Large, prospective, multicenter studies following well-defined inclusion and exclusion criteria would bypass the single-institution, small-size limitation and would be helpful for future studies.

    Conclusions

    In this cohort study of patients with FMF, cutaneous stage was the most effective variable associated with survival among all variables tested. We subanalyzed overall survival and DSS by cutaneous stage, finding meaningful clinical differences, and this is the first US study, to our knowledge, to demonstrate this difference. Therefore, we hold that this system is helpful and advocate its use for patients with FMF. Furthermore, our data support that FMF is not intrinsically an aggressive mycosis fungoides variant because patients with early-stage cutaneous disease have a prognosis essentially equivalent to that of early-stage CMF. However, a subset of patients with advanced cutaneous stage FMF have an aggressive course. This cutaneous staging scheme is of practical utility in the clinical setting because it will give patients with FMF a more realistic sense of their prognosis.

    Back to top
    Article Information

    Accepted for Publication: September 12, 2020.

    Published Online: December 9, 2020. doi:10.1001/jamadermatol.2020.4372

    Corresponding Author: Laura B. Pincus, MD, Dermatopathology Service, Departments of Pathology and Dermatology, University of California, San Francisco, 1701 Divisadero St, Ste 280, San Francisco, CA 94115 (laura.pincus@ucsf.edu).

    Author Contributions: Drs Charli-Joseph and Pincus had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Charli-Joseph, Allen, Pincus.

    Critical revision of the manuscript for important intellectual content: Charli-Joseph, Kashani-Sabet, McCalmont, Kornak, Ai, LeBoit, Pincus.

    Statistical analysis: Charli-Joseph, Kornak, Allen, Pincus.

    Obtained funding: Pincus.

    Administrative, technical, or material support: McCalmont, LeBoit, Pincus.

    Supervision: Kashani-Sabet, McCalmont, LeBoit, Pincus.

    Conflict of Interest Disclosures: Dr Pincus reported receiving grants from the Dermatology Foundation during the conduct of the study. No other disclosures were reported.

    Funding/Support: This study was supported by grant A112139 from the Dermatology Foundation (Dr Pincus).

    Role of the Funder/Sponsor: The Dermatology Foundation 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.

    Additional Contributions: We thank the patients for granting permission to publish this information. Chengshi Jin, PhD, MS, Department of Epidemiology and Biostatistics, University of California, San Francisco, provided statistical support, and John G. Williams, MEng(hons), Harris Design Studio, gave technical advice on image processing. Neither was compensated for this work.

    References
    1.
    Willemze  R, Cerroni  L, Kempf  W,  et al.  The 2018 update of the WHO-EORTC classification for primary cutaneous lymphomas.   Blood. 2019;133(16):1703-1714. doi:10.1182/blood-2018-11-881268 PubMedGoogle ScholarCrossref
    2.
    Swerdlow  SH, Campo  E, Harris  NL,  et al.  World Health Organization Classification of Tumours of Hematopoietic and Lymphoid Tissues. IARC Press; 2008.
    3.
    Swerdlow  SH, Campo  E, Pileri  SA,  et al.  The 2016 revision of the World Health Organization classification of lymphoid neoplasms.   Blood. 2016;127(20):2375-2390. doi:10.1182/blood-2016-01-643569 PubMedGoogle ScholarCrossref
    4.
    van Doorn  R, Scheffer  E, Willemze  R.  Follicular mycosis fungoides, a distinct disease entity with or without associated follicular mucinosis: a clinicopathologic and follow-up study of 51 patients.   Arch Dermatol. 2002;138(2):191-198. doi:10.1001/archderm.138.2.191 PubMedGoogle ScholarCrossref
    5.
    Gerami  P, Rosen  S, Kuzel  T, Boone  SL, Guitart  J.  Folliculotropic mycosis fungoides: an aggressive variant of cutaneous T-cell lymphoma.   Arch Dermatol. 2008;144(6):738-746. doi:10.1001/archderm.144.6.738 PubMedGoogle ScholarCrossref
    6.
    Lehman  JS, Cook-Norris  RH, Weed  BR,  et al.  Folliculotropic mycosis fungoides: single-center study and systematic review.   Arch Dermatol. 2010;146(6):607-613. doi:10.1001/archdermatol.2010.101 PubMedGoogle ScholarCrossref
    7.
    Mantaka  P, Helsing  P, Gjersvik  P, Bassarova  A, Clausen  OP, Delabie  J.  Clinical and histopathological features of folliculotropic mycosis fungoides: a Norwegian patient series.   Acta Derm Venereol. 2013;93(3):325-329. doi:10.2340/00015555-1476 PubMedGoogle ScholarCrossref
    8.
    Demirkesen  C, Esirgen  G, Engin  B, Songur  A, Oğuz  O.  The clinical features and histopathologic patterns of folliculotropic mycosis fungoides in a series of 38 cases.   J Cutan Pathol. 2015;42(1):22-31. doi:10.1111/cup.12423 PubMedGoogle ScholarCrossref
    9.
    Muniesa  C, Estrach  T, Pujol  RM,  et al.  Folliculotropic mycosis fungoides: clinicopathological features and outcome in a series of 20 cases.   J Am Acad Dermatol. 2010;62(3):418-426. doi:10.1016/j.jaad.2009.03.014 PubMedGoogle ScholarCrossref
    10.
    Hodak  E, Amitay-Laish  I, Feinmesser  M,  et al.  Juvenile mycosis fungoides: cutaneous T-cell lymphoma with frequent follicular involvement.   J Am Acad Dermatol. 2014;70(6):993-1001. doi:10.1016/j.jaad.2013.12.029 PubMedGoogle ScholarCrossref
    11.
    Marschalkó  M, Erős  N, Kontár  O,  et al.  Folliculotropic mycosis fungoides: clinicopathological analysis of 17 patients.   J Eur Acad Dermatol Venereol. 2015;29(5):964-972. doi:10.1111/jdv.12743 PubMedGoogle ScholarCrossref
    12.
    Tomasini  C, Kempf  W, Novelli  M,  et al.  Spiky follicular mycosis fungoides: a clinicopathologic study of 8 cases.   J Cutan Pathol. 2015;42(3):164-172. doi:10.1111/cup.12399 PubMedGoogle ScholarCrossref
    13.
    Hodak  E, Amitay-Laish  I, Atzmony  L,  et al.  New insights into folliculotropic mycosis fungoides (FMF): a single-center experience.   J Am Acad Dermatol. 2016;75(2):347-355. doi:10.1016/j.jaad.2016.03.009 PubMedGoogle ScholarCrossref
    14.
    Deonizio  JM, Ascef  RD, Sanches  JA.  Folliculotropic mycosis fungoides: clinical and epidemiological evaluation in a single center in Brazil.   Int J Dermatol. 2016;55(5):e256-e261. doi:10.1111/ijd.13149 PubMedGoogle ScholarCrossref
    15.
    van Santen  S, Roach  RE, 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 PubMedGoogle ScholarCrossref
    16.
    Kim  YH, Liu  HL, Mraz-Gernhard  S, Varghese  A, Hoppe  RT.  Long-term outcome of 525 patients with mycosis fungoides and Sezary syndrome: clinical prognostic factors and risk for disease progression.   Arch Dermatol. 2003;139(7):857-866. doi:10.1001/archderm.139.7.857 PubMedGoogle ScholarCrossref
    17.
    Agar  NS, Wedgeworth  E, Crichton  S,  et al.  Survival outcomes and prognostic factors in mycosis fungoides/Sézary syndrome: validation of the revised International Society for Cutaneous Lymphomas/European Organisation for Research and Treatment of Cancer staging proposal.   J Clin Oncol. 2010;28(31):4730-4739. doi:10.1200/JCO.2009.27.7665 PubMedGoogle ScholarCrossref
    18.
    Olsen  E, Vonderheid  E, Pimpinelli  N,  et al; ISCL/EORTC.  Revisions to the staging and classification of mycosis fungoides and Sezary syndrome: a proposal of the International Society for Cutaneous Lymphomas (ISCL) and the cutaneous lymphoma task force of the European Organization of Research and Treatment of Cancer (EORTC).   Blood. 2007;110(6):1713-1722. doi:10.1182/blood-2007-03-055749 PubMedGoogle ScholarCrossref
    19.
    Cerroni  L, Fink-Puches  R, Bäck  B, Kerl  H.  Follicular mucinosis: a critical reappraisal of clinicopathologic features and association with mycosis fungoides and Sézary syndrome.   Arch Dermatol. 2002;138(2):182-189. doi:10.1001/archderm.138.2.182 PubMedGoogle ScholarCrossref
    ×