Phototherapy of mycosis fungoides
Conflicts of interest:
Franz Trautinger, Department of Dermatology and Venereology, Landesklinikum St. Pölten, Propst-Fuehrer-Strasse, St Pölten A-3100, Austria
Tel: +43 2742 300 11909
Fax: +43 2742 300 11919
Background/purpose: Among the primary cutaneous T-cell lymphomas, mycosis fungoides (MF) is the most common disease entity. Recently, an improved understanding of the pathology, clinical presentation, and prognosis of MF has lead to the development of new and practically useful classification and staging systems. In most patients, MF presents with patches and plaques and remains confined to the skin for years and decades, making it an ideal target for phototherapy. However, treatment schedules vary widely and this review describes the current knowledge about phototherapy of MF focusing mainly on narrow- and broadband UVB and 8-methoxypsoralen plus UVA, its indications, practical aspects, and clinical outcome.
Methods: Review and summary of the pertinent literature.
Results and conclusions: Since 1976, when the first report on phototherapy for MF was published, sufficient evidence has accumulated to make narrowband UVB and PUVA safe and effective treatment options for early stages of the disease. In refractory cases or more advanced stages, combination of phototherapy with systemic treatments including mainly interferons and retinoids might be valuable. Additional research is required to further define the optimal treatment schedules and the role of maintenance.
Mycosis fungoides (MF): epidemiology, classification, staging, and prognosis
Cutaneous T-cell lymphomas (CTCL) are a heterogenous group of non-Hodgkin lymphomas (NHL), which are characterized by their initial presentation in the skin. As with other NHL, major progress has been made recently with the development of classification systems that aim to define disease entities based on pathology, clinical presentation, and prognosis (1, 2). MF (MF) is the most common form (about 65%) of CTCL. Its age-adjusted incidence rate in the United States has been recently estimated to be 4.1/1 000 000 with a male predominance (1, 3, 4). Sezary syndrome (SS) – although usually mentioned together with MF – is a separate entity according to the European Organization for the Research and Treatment of Cancer (EORTC)-WHO classifications (1, 2). SS is exceedingly rare (4% of CTCL) and – although sharing histopathological features – has a different clinical presentation and a worse prognosis than MF (1, 5). SS will not be mentioned here in more detail since skin-directed therapies (SDT) – if at all – play a minor role in its treatment.
Histopathologically, MF is characterized by an epidermotropic infiltrate of small- to medium-sized T lymphocytes with cerebriform nuclei. Typically, the lymphocytes display the immunophenotype of mature memory T-cells (CD3+, CD4+, CD45RO+, CD8−) and clonal rearrangement of the T-cell receptor gene is detectable in most cases. Clinically, the disease presents with erythematous, sometimes poikilodermatous and scaly patches and plaques with an occasional itch (1, 6). MF has an indolent and chronic clinical course characterized by episodes of treatment-associated remissions and subsequent relapses. In patients with early stages (stages T1–2, N0–2, M0, B0, 1 are usually described as early stages in the literature, see Table 1), life expectancy is not significantly different from age-matched controls. Progression to extracutaneous involvement is estimated to occur in up to 30% of patients and is associated with a poor prognosis (5). There is insufficient evidence to date to conclude that the natural course of the disease could be modified by any type of therapy (with the exception of hematopoietic cell transplantation in selected patients) and thus palliation is the major goal of treatment (7, 8). Consequently, ‘expectant policy’ including careful observation without active treatment is considered a valid option for some patients with early disease stages (8).
Table 1. Tumor-node-metastasis-blood (TNMB) staging as proposed by the ISCL/EORTC (10)
|T1||Limited patches, papules, and/or plaques covering <10% of skin surface|
|T2||Patches, papules, and/or plaques covering ≥10% of skin surface|
|T3||One or more tumors (≥1 cm in diameter)|
|T4||Confluence of erythema covering ≥80% body surface area|
|N0||No clinically abnormal peripheral lymph nodes, biopsy not required|
|N1a||Clinically abnormal peripheral lymph nodes, histopathology: dermatopathic lymphadenopathy, clone negative|
|N1b||Clinically abnormal peripheral lymph nodes, histopathology: dermatopathic lymphadenopathy, clone positive|
|N2a||Clinically abnormal peripheral lymph nodes, histopathology: early involvement of MF, clone negative|
|N2b||Clinically abnormal peripheral lymph nodes, histopathology: early involvement of MF, clone positive|
|N3||Clinically abnormal peripheral lymph nodes, histopathology: partial or complete effacement of lymph node architecture|
|Nx||Clinically abnormal peripheral lymph nodes, no histologic confirmation|
|M0||No visceral organ involvement|
|B0a||Absence of significant blood involvement, clone negative|
|B0b||Absence of significant blood involvement, clone positive|
|B1a||Low tumor burden: >5% of peripheral blood lymphocytes are atypical cells but does not meet the criteria of B2, clone negative|
|B1b||Low blood tumor burden: >5% of peripheral blood lymphocytes are atypical cells but does not meet the criteria of B2, clone positive|
|B2||High blood tumor burden: ≤1000/μl Sezary cells with positive clone|
The first staging system for MF has been published in 1979 by Bunn and Lamberg (9). Although never formally validated this system proved to be clinically useful and until recently remained the standard for clinical studies and patient care. Based on advances in our knowledge and understanding of the pathology, molecular biology, classification, and clinical course of the disease, a modified staging system was proposed by the International Society for Cutaneous Lymphomas and the Cutaneous Lymphoma Task Force of the EORTC (Tables 1 and 2) (10). Recently, this system has been validated in a cohort of 1502 patients from the United Kingdom (5).
Table 2. Staging classification based on the updated tumor-node-metastasis-blood (TNMB) staging as proposed by the ISCL/EORTC (10)
The role of phototherapy in the management of MF
There is a lack of randomized clinical trials in MF and thus the choice of treatment is often determined by availability, experience, and preference of both, physician and patient. Recently, however, treatment recommendations and reviews have been published that try to summarize the available evidence and provide a more rational approach to the management of MF (8, 11–13). Based on these publications and on the results of an early seminal study, it is now commonly accepted that early-stage MF should be initially treated with SDT (14). Systemic and more aggressive treatments should be reserved for higher stages (≥IIB), progression, or lack of appropriate responses. Irrespective of disease stage and type of treatment, maintenance of quality of life should be at the center of therapeutic strategies. SDT that have shown clinical efficacy in MF include topical corticosteroids, topical chemotherapy (nitrogen mustard, carmustine), topical bexaroten (not available in Europe), radiotherapy (total skin electron beam, superficial X-irradiation), and phototherapy (8).
Remarkably, MF (together with other types of CTCL) is the only malignant disease that is treated with ultraviolet (UV) radiation, the major environmental skin carcinogen. The role of UV in the pathogenesis of MF is unclear. On the one hand, epidemiologic studies indicate that exposure to UV is associated with a decreased risk for the development of NHL without allowing for a reliable conclusion regarding NHL subtypes, including MF (15, 16). On the other hand, McGregor et al. (17) have described UV-specific p53 mutations in advanced MF, pointing to a role for ultraviolet B (UVB) in the pathogenesis and progression of MF.
The first publication on phototherapy for MF appeared in 1976 when Gilchrest et al. (18) reported on the treatment of nine patients with 8-methoxypsoralen plus UVA (PUVA). All patients responded favorably and complete remission was achieved in four patients. Through shielding of one arm during the initial treatment phase, the authors clearly demonstrated the specific efficacy of the treatment and the absence of a relevant systemic effect of PUVA. Since then, new light sources and treatment modalities have been developed, and in addition to PUVA, standard phototherapy options for MF now include narrow- and broadband UVB. Extracorporeal photochemotherapy (photopheresis, ECP) has been developed for the treatment of SS and MF with erythroderma (19). During ECP, the patients' peripheral blood leukocytes are exposed to 8-methoxypsoralen and UVA. Comprehensive reviews on ECP have been published recently and the treatment will not be covered here (20, 21). Other modalities where only small case series have been published to date include photodynamic therapy, excimer laser, and UVA1 (22–29). These treatments have not found widespread application in MF because they are either limited to the irradiation of very limited skin areas, or the equipment is expensive and not widely available, or both.
As mentioned above, the first report on the use of PUVA for MF appeared in 1976 (18). The authors concluded that ‘methoxsalen photochemotherapy may prove to be a valuable addition to therapies currently available for MF and may obviate some of the problems associated with conventional management of this disorder’. Although >30 years have passed and although in the meantime a variety of other treatment forms for MF have been introduced, PUVA has still retained its value in the treatment of MF. This is supported by the results from a large number of reports, which – taken together – might allow for a reliable estimate of response rates. A comprehensive list of these studies has been recently published (30). As is often the case with studies on diseases that are rare and run a chronic course, the available data suffer from a wide heterogeneity regarding patient selection, treatment protocols, outcome measures, observation periods, concomitant treatments, sample size, and general quality. However, based on the data of five studies and a total of 244 patients, Hermann and colleagues have calculated the rate of complete remission after an initial course of PUVA to be 90% for IA, 76% for stage IB, 78% for stage IIA, 59% for stage IIB%, and 61% for stage III (staging according to (9)) and similar figures have been reported by others (31, 32).
Less well investigated, however, is the long-term outcome regarding relapse rates and disease-free survival. In a follow-up study of 44 patients, Hönigsmann et al. (33) reported that five of nine (stage IA) and 10 of 26 (stage IB) patients remained in remission for an observation period of up to 79 months. All patients with T3 disease experienced relapses. Mean disease-free survival was 20 months for stage IA and 17 months for stage IB. In a more recent study, 66 patients with stages IA to II who achieved complete remission after initial PUVA were followed up for up to 242 months (34). 5- and 10-year disease-free survival rates for patients with T1 disease were reported to be 56% and 30%, respectively, and 74% and 50% for T2. Actuarial survival rates at 5, 10, and 15 years did not differ between the nonrelapse and relapse groups.
In summary, there is good evidence that PUVA is effective for the clearing of skin lesions, particularly in early stages of the disease, and that long-term remissions can be achieved in a certain percentage of patients. How PUVA in this indication compares with other treatment modalities has not been formally investigated. Retrospective data published for topical chemotherapy and total skin electron beam radiation show very similar outcomes as described above for PUVA (35–37). Thus, as long as no data from comparative studies are available, the choice of treatment in early MF can be made on availability, experience of the physician, and preference of the patient.
Important open questions regarding the use of PUVA in MF relate to the duration of treatment, dose escalation, maintenance therapy, combination therapies, and short- and long-term toxicity. Technically, PUVA in MF is generally performed according to the guidelines established for the treatment of psoriasis (38, 39). Although bath PUVA with 8-methoxypsoralen has been shown in a retrospective analysis to be effective, its use in MF is not generally accepted because the head is usually not exposed to the photosensitizer and relapses might preferentially occur in nontreated areas (40). Treatment is usually applied until complete clinical clearing is achieved. A confirmatory biopsy of a previously exposed site is often recommended at this point, although the implication of clinical remission without pathological clearance is unclear. Thus, a more reasonable approach seems to be to biopsy only lesions, which are clinically equivocal to guide further treatment decisions.
In a recent multinational survey among dermatologists, 88% of the respondents indicated that they use some form of PUVA maintenance after clearance has been achieved (30). This result is in line with the published practice, because most studies included some form of maintenance. There is, however, no agreement on duration, frequency, and UVA-dose and a wide range of schedules are in use. In summary, there is insufficient evidence for the benefit of PUVA maintenance. In the light of the carcinogenic risk associated with PUVA, which has been clearly documented in patients with psoriasis, and the consequent recommendation to minimize exposure, the recently published consensus of the EORTC suggested to avoid maintenance (8, 41–43). A practical approach might be to reserve maintenance for patients who experience an early relapse after an initial course of PUVA and to adapt the schedule as to minimize interference with the patients' quality of life (e.g. once weekly without further dose increments for 3–6 months).
Through combination with systemic treatments, it is aimed to enhance the efficacy of PUVA in later disease stages or in patients with insufficient response to PUVA alone. The most commonly used combination partners are interferon-α (IFN-α) and retinoids (isotretinoin, etretinate, acitretin). These drugs have shown some efficacy as monotherapy in the treatment of MF and through combination with PUVA an additive effect seemed likely (44–49). More recently, bexarotene, a new retinoid that binds to the nuclear retinoid × receptor (R×R), has been specifically developed for the treatment of MF (50). The drug has shown efficacy as monotherapy in randomized clinical trials and an increasing number of case series and smaller studies indicate that its combination with PUVA might be safe and feasible (51–58). In general, however, the questions as to whether any PUVA combination is superior to PUVA alone in terms of clinically relevant endpoints (e.g. overall survival, disease-free survival, toxicity) remain unsolved. Studies are either retrospective, too small, lacking appropriate controls, or used surrogate endpoints (e.g. duration of treatment or total UVA dose). A randomized trial that compared acitretin combined with either PUVA or IFN-α in patients with stages I and II found out that the PUVA combination is more efficient in inducing remissions after at least 4 weeks of treatment (59). Information on remission-free periods and long-term outcomes were not reported. Thus, because it is still unclear whether any combination is superior to monotherapy, it is a reasonable recommendation to start with phototherapy alone and combine it with one of the above mentioned drugs (depending on individual contraindications and tolerability) in case of insufficient response or early relapse. For patients with higher stages of the disease (e.g. tumor stage or lymph node involvement), combination of PUVA with a systemic treatment might be a primary choice to provide palliation of the skin symptoms at the same time with control of the systemic involvement.
Acute side effects of PUVA include nausea, pruritus, and phototoxic reactions, presenting as painful erythema, edema, and sometimes blistering. These reactions do not differ between MF and other skin diseases treated with PUVA. Sometimes, MF-lesions appear more pronounced during the initial treatment phase, which might be mistakenly attributed to overdose. Regarding long-term safety, PUVA has been shown to be a carcinogen and treatment is dose dependently associated with the risk of squamous cell carcinomas. This has been convincingly demonstrated in a large cohort of patients with psoriasis (43). Although similar studies are not available for MF, the pertinent literature indicates that there is also a similar risk in these patients for the development of nonmelanoma skin cancer (60). This has to be taken into account when PUVA is considered and when obtaining a patient's informed consent.
With the development of fluorescent light bulbs that emit in the UVB range, phototherapy with these lamps was introduced initially mainly for psoriasis and vitiligo (61). The treatment had the advantages of being easy to use and suitable for outpatient or even home treatment. The first report of UVB-phototherapy for MF appeared in 1982 (62). In this study, 31 patients were treated and in 61% reportedly complete remissions were achieved. In a long-term follow-up of the same patient cohort, it was reported that 23% of the patients remained in remission for prolonged periods even, after discontinuation of maintenance (63). Because this initial observation, similar results have been reported in a number of other studies. A comprehensive list and summary of these studies have been recently published elsewhere (32). A technical innovation that specifically changed phototherapy was the development of the TL-01 fluorescent lamp (64). This lamp emits within a narrow spectrum of around 311 nm and thus according to the available action spectra should be ideally suitable for the clearance of psoriatic lesions with a minimized erythema risk (65). Accordingly, narrowband UVB (NB-UVB) has been shown to be better than broadband UVB and similarly effective as PUVA in the treatment of psoriasis (66). As a consequence to date, TL-01 lamps have replaced broadband UVB sources in many phototherapy wards. According to a recent survey among dermatologists, who use office-based phototherapy to treat patients with MF, NB-UVB turned out to be first choice for T1 disease in phototypes I and II (30). In this clinical situation, two out of 26 respondents only recommended broadband UVB. In later stages or darker skin phototypes, PUVA took over as first choice, but again broadband UVB was recommended by almost none of the dermatologists participating in the survey. The efficacy of NB-UVB has been addressed by a number of studies and case series. In a study by Hofer et al. (67), complete remissions in 19 out of 20 patients has been achieved by a treatment course of 5–10 weeks with three to four exposures per week. Similar remission rates have been published by others and from the cumulated data it can be concluded, that NB-UVB – as the other modalities described above – can induce high rates of complete remissions in early MF (30, 32). Evidence is poor, however, regarding the question how NB-UVB compares with other treatments. Diederen et al (68) published a retrospective analysis of 56 patients treated with NB-UVB (n=21) or PUVA (n=35). Complete remission rates were similar (81% for NB-UVB, 71% for PUVA) and mean relapse-free intervals were almost identical (24.5 and 22.8 months, respectively). The authors concluded that due to the practical advantages of NB-UVB it might be a reasonable approach to a patient with early MF to start with NB-UVB and in case of lack of progression or lack of response switch to PUVA. Similarly, Ahmad et al. (69) reported in 2007 from a retrospective analysis of the records of 40 patients that both, PUVA and NB-UVB, are effective in the treatment of early MF.
Acute toxicity from NB-UVB is almost negligible with only mild erythema that usually fades within 24 h. Although it is likely that NB-UVB is associated with a risk for skin cancer, this has not yet been quantified (70) and is probably well below that of PUVA.
Areas of uncertainty
After >30 years of experience in the treatment of MF, it is evident that the various forms of phototherapy are potent tools to clear superficial skin lesions in early MF. It is, however, still unclear whether these treatment are able to modify the natural course of the disease. Evidence from sufficiently powered, randomized, controlled trials on the efficacy of phototherapy in terms of the clinically most relevant endpoints, namely disease-free survival and long-term safety, is still lacking (overall survival is almost uncompromised in early MF and thus can not be reasonably used as an outcome measure in these patients). The newly developed classification and staging systems for the first time provide the basis that such studies can be performed on well-defined target populations. Furthermore, there are insufficient data to conclude based on a higher level of evidence which type of phototherapy and which schedule (e.g. in terms of dose increments, frequency, etc.) might be optimal. The same is true for maintenance treatments, for which – although widely used in clinical practice – efficacy has not yet been proven (71).
Conclusions and recommendations
Phototherapy with (NB-)UVB and PUVA are useful and highly effective in inducing remission of patches and plaques in patients with MF. Treatment should be used according to the guidelines for phototherapy as published by various national and international dermatological societies (39, 70). The choice of treatment largely depends on availability, patient history (in terms of earlier treatment responses), and experience of the physician. It has become a widely accepted consensus that in patients with patches and thin plaques NB-UVB should be preferentially used and that PUVA should be reserved for patients with thick plaques (including follicular mucinosis), with photoypes ≥III, and insufficient response to UVB (8).
Our understanding of the disease has significantly increased recently and ongoing research is likely to yield new insights into its pathology as well as new forms of treatment (72). It is, however, not likely that phototherapy of MF will become obsolete in the foreseeable future, because new therapeutic developments will rather supplement than replace it, as can be seen from the example of bexarotene.