Mycosis fungoides (MF) is a low-grade, non-Hodgkin’s lymphoma caused by skin-homing CD4+ T cells that form cutaneous patches, plaques, and tumors.^1,² MF was initially noted in 1806 when Alibert described a patient with cutaneous tumors that he attributed to yaws. Although the disease was initially termed pian fungoides, he later changed the name to mycosis fungoides.³ In 1938, Sézary and Bouvrain described a leukemic variant called Sézary syndrome (SS),⁴ and Lutzner and Jordan elucidated the ultrastructure of the Sézary cell in 1968.⁵ The term cutaneous T-cell lymphoma was introduced by Edelson in 1975 and encompasses a variety of cutaneous lymphoproliferative disorders, including MF/SS, adult T-cell leukemia/lymphoma, primary cutaneous CD30+ anaplastic lymphoma, lymphomatoid papulosis, pagetoid reticulosis, and others.⁶ In clinical practice, the terms MF and cutaneous T-cell lymphoma (CTCL) are often used interchangeably; however, such usage is incorrect.⁷ MF constitutes only 50% of all CTCLs, and the clinical history and therapy for each subtype of CTCL are different.²

MF is a challenging disorder from all perspectives. Despite improving molecular techniques, diagnosis early in the course of disease is often difficult because of the nonspecific nature of skin lesions and the numerous benign dermatoses that may mimic MF. Once a diagnosis of MF has been correctly established, the optimal initial treatment strategy often remains unclear, given the heterogeneity of clinical presentations and limited data from controlled studies. Although radiotherapy is the most effective single agent in the treatment of MF,⁸ total skin electron beam therapy (TSEBT) is not readily available at many centers. This chapter provides a summary of the clinically relevant aspects of MF and describes the role and techniques of radiotherapy in patient management.

Etiology and Epidemiology

MF primarily affects adults over the age of 40 years, with incidence rates peaking in the seventh decade. The incidence of CTCL has consistently increased from 1974 through 2002, reaching a current yearly incidence rate of 9.6 cases per million.⁹ Risk factors for the development of CTCL include black race and male gender. Markers of high socioeconomic status, such as residence in areas with high home values, high level of educational attainment, and high physician density, are associated with an increased incidence CTCL; it is unclear whether these factors are causative or simply increase the likelihood of diagnosis.⁹

Etiologic agents for the development of MF remain highly speculative.¹⁰ Although numerous exposures to toxins, including pesticides, radiation, industrial solvents, tobacco, and alcohol, have been investigated, no consistent causative factors have been identified.¹¹ High rates of seropositivity for both cytomegalovirus¹² and human T-cell lymphotrophic virus type I^13,¹⁴ have been reported, but these studies await further corroboration before a causal relationship can be inferred.^15,16,17

The observation that MF is more common in blacks and tends to present in sun-shielded areas (i.e., bathing suit distribution) suggests that sun exposure may protect against the development of MF. Sun exposure may mediate its protective effect by exerting a cytotoxic effect on either the malignant CD4+ T cell of MF or the epidermal antigen-presenting dendritic cell, also known as the Langerhans cell. This cell presents antigens to the malignant CD4+ T cells of MF and may stimulate their growth. The histologic evidence for this interaction is Pautrier’s microabscess, an intraepidermal collection of malignant CD4+ T cells clustered around an antigen-presenting dendritic cell. This finding is considered pathognomonic for MF and suggests that MF may be an antigen-driven malignant disease, though a specific antigen has yet to be identified.¹⁸

Prevention and Early Detection

No agents have been identified that will prevent the development of MF. Early diagnosis is critical, however, because local therapy directed against unilesional or oligolesional MF is highly curative.¹⁹^–²¹ The most typical presentation of early disease—an erythematous patch with scale arising in a sun-shielded area—may be confused with a number of benign dermatoses, including atopic dermatitis, psoriasis, and tinea corporis.²² At this early stage, most of the lymphocytes noted on histopathologic studies represent reactive inflammatory cells rather than the malignant clone.²³ As a result, the histopathologic findings of early MF mimic numerous benign inflammatory conditions,^23,²⁴ and a rapid and correct histologic diagnosis is not always possible. In this setting, molecular studies such as the polymerase chain reaction (PCR) for the T-cell receptor will identify a clonal T-cell population in 50% to 80% of patients who ultimately develop overt histologic evidence of MF.^25,²⁶ To improve the diagnostic accuracy of early MF, the International Society for Cutaneous Lymphoma proposed a points-based algorithm for early diagnosis²⁷ (Table 77-1).

TABLE 77-1 Proposed Algorithm for Diagnosis of Early Mycosis Fungoides

Criteria	Score
Clinical	2 points for basic + two additional criteria 1 point for basic + one additional criterion
Basic
Persistent and/or progressive patches/thin plaques
Additional
1. Non–sun-exposed location
2. Size/shape variation
3. Poikiloderma*
Histopathologic	2 points for basic + two additional criteria 1 point for basic + one additional criterion
Basic
Superficial lymphoid infiltrate
Additional
1.Epidermotropism without spongiosis
2.Lymphoid atypia†
Molecular Biologic	1 point for clonality
Basic
Clonal T-cell receptor gene rearrangement
Immunopathologic	1 point for one or more criteria
1.<50% CD2+, CD3+, and/or CD5+ T cells
2.<10% CD7+ T cells
3.Epidermal/dermal discordance of CD2, CD3, CD5, or CD7‡
Four or more points satisfy criteria for a diagnosis of early MF.

* Poikiloderma is defined as the combination of skin atrophy, telangiectasia, and mottled pigmentation.

† Lymphoid atypia is defined as cells with enlarged, hyperchromatic nuclei and irregular or cerebriform nuclear contours.

‡ T-cell antigen deficiency confined to the epidermis.

Adapted from Pimpinelli N, Olsen EA, Santucci M, et al: Defining early mycosis fungoides. J Am Acad Dermatol 53:1053-1063, 2005.

Biologic Characteristics and Molecular Biology

A malignant clone in patch-plaque MF bears the immunophenotype of activated, skin-homing CD4+ helper T cells.¹⁸ When a naïve T cell identifies its cognate antigen in a skin-draining lymph node, activation occurs, and the T cell begins to express cutaneous lymphocyte antigen (CLA) and CC chemokine receptor 4 (CCR4). As these activated T cells pass through the capillaries of inflamed skin, CLA and CCR4 bind to their respective ligands on the dermal capillaries, resulting in extravasation of the activated T cells into the dermal connective tissue. Once outside the circulation, activated T cells migrate to the epidermis and interact with antigen-presenting dendritic (Langerhans) cells (Fig. 77-1).

Figure 77-1 Molecular pathology of mycosis fungoides . Activated skin-homing T cells extravasate through the dermal capillaries due to interactions of cutaneous lymphocyte antigen (CLA) and chemokine receptor 4 (CCR4) with their respective ligands on the dermal capillaries, E-selectin, and chemokine ligand 17 (CCL17). T cells then migrate to the epidermis and interact with antigen-presenting dendritic cells.

From Girardi M, Heald PW, Wilson LD: Medical progress. The pathogenesis of mycosis fungoides. N Engl J Med 350:1978-1988, 2004.

Clinically, progression of MF is associated with loss of epidermotropism and increasing tumor burden. Molecular studies have shown that progression of MF is associated with TP53 mutation,²⁸ numerous chromosomal rearrangements,²⁹ and microsatellite instability.³⁰ In addition, the malignant cells in MF develop mechanisms to escape destruction by the host immune system. For example, although benign activated T cells are eliminated by fas/fas-ligand-mediated apoptosis, the malignant T cells of MF evade fas-mediated apoptosis via fas downregulation, mutation, or alternative splicing.³¹

As a malignant disease of the immune system, MF results in substantial alteration of host immunity with consequent increased risk of infection ³² and possibly secondary malignant disease.³³ For example, in patients with Sézary syndrome, the absolute number of normal circulating T cells often drops dramatically, reaching levels typically seen only in acquired immunodeficiency syndrome.³⁴ In addition, malignant CD4+ T cells produce large amounts of interleukin-10 and transforming growth factor-β, resulting in further suppression of cell-mediated immunity.^35,³⁶ The malignant cells of MF also produce large amounts of soluble interleukin-2 receptor, which can inactivate interleukin-2,³⁷ a cytokine needed to promote normal T-cell activation. Finally, the malignant cells of Sézary syndrome can elaborate large amounts of interleukin-4 and interleukin-5, producing a syndrome characterized by atopy and eosinophilia.³⁶

Pathology and Pathways of Spread

The diagnosis of MF remains challenging, even for the experienced clinician and dermatopathologist, due to both the absence of a diagnostic gold standard and the number of benign inflammatory dermatoses that may mimic MF, particularly in its early stages. Currently, the diagnosis relies upon integrating the clinical presentation with the histopathologic, immunophenotypic, and genotypic data. The current World Health Organization–European Organization for the Research and Treatment of Cancer (WHO-EORTC)³⁸ pathologic classification scheme for CTCL is presented in Table 77-2.

TABLE 77-2 WHO-EORTC Classification for Cutaneous T-Cell Lymphomas

Histopathology

The most striking finding of early MF is profound epidermotropism, characterized by lymphocytes clustered along the basement membrane of the epidermis ³⁹ (Fig. 77-2A and B). Microdissection studies have shown that virtually all of the lymphocytes in the epidermis belong to the malignant clone, whereas most dermal lymphocytes are reactive.^40,⁴¹

Figure 77-2 Classic histopathologic findings in mycosis fungoides and Sézary syndrome. A, Advanced patch-stage lesion of mycosis fungoides (MF) exhibiting abundant lymphocytes within the basal layer of the epidermis, associated with an underlying bandlike lymphocytic infiltrate and papillary dermal fibrosis. B, Advanced patch-stage lesion of MF exhibiting enlarged, convoluted lymphocytes within the epidermis. The lymphocyte cell size approximates the width of keratinocyte nuclei. C, Plaque-stage MF shows, in addition, involvement of the reticular dermis. D, Pautrier’s microabscesses are well-defined aggregates of lymphocytes within the epidermis, which are strongly indicative of MF. E, In tumor-stage MF, the dermis is distended by the lymphocytic infiltrate. Epidermotropic capacity is often lost. F, Sézary cells, such as the cell at the right, are enlarged circulating lymphocytes that exhibit cerebriform nuclear convolutions.

Photos and legend courtesy Earl Glusac, MD; reprinted with permission from Figure 5 in Smith BD, Wilson LD: Management of mycosis fungoides. Part 1. Diagnosis, staging, and prognosis. Oncology 17:1281-1288, 2003.

Several microscopic findings help to discriminate early MF from benign inflammatory mimics. For example, an EORTC study reported that identification of epidermal lymphocytes with extremely convoluted, medium to large (7 to 9 µm) nuclei enabled a correct diagnosis of MF with 100% sensitivity and 92% specificity ²⁴ (see Fig. 77-2B). In contrast, a study from Stanford University found that intraepidermal atypical lymphocytes surrounded by a clear halo (an artifact of fixation) were the most robust indicator of MF in a multivariate model.⁴² Another finding, Pautrier’s microabscess, is considered pathognomonic but is seen in less than 20% of early lesions⁴² (see Fig. 77-2D). Efforts are currently under way to establish a grading system to standardize pathology reporting and improve diagnostic accuracy.^23,⁴³

As MF progresses from the patch to plaque stage, the lymphoid infiltrate increases in density and begins to invade the deeper reticular dermis (see Fig. 77-2C). Furthermore, due to the increased burden of neoplastic cells, findings such as Pautrier’s microabscesses (see Fig. 77-2D), haloed lymphocytes, and convoluted nuclei are more readily identified, resulting in improved diagnostic accuracy. Tumor formation results from vertical growth of the lymphoid infiltrate and may be associated with complete loss of epidermotropism and sparing of the upper papillary dermis (see Fig. 77-2E). Erythrodermic MF often resembles patch-stage MF, although epidermotropism may be more subtle and neoplastic cells may be quite sparse.⁴⁴

Immunophenotyping of Skin Lesions

Assessment of T-cell marker expression within the lymphoid infiltrate provides additional information that helps to establish a diagnosis of MF. The hallmark of MF is expression of CD4, the marker of mature helper T cells. A typical immunophenotype for MF is CD2+ (pan T cell), CD3+ (pan T cell), CD4+ (helper T cell), CD5+ (pan T cell), CD45RO+ (memory T cell), CLA+ (cutaneous lymphoid antigen), CD8− (cytotoxic T cell), CD30− (activated T cell).⁴⁵ Although many benign dermatoses express a similar immunophenotype, two markers, CD7 and Leu-8, are often underexpressed in MF and may be helpful in distinguishing between MF and benign mimics. Finally, rare cases of apparently classic MF that are CD4− but CD3+ and CD8+ have been reported.⁴⁶

T-Cell Receptor Gene Rearrangement

Clonal T-cell receptor gene rearrangements are frequently identified in MF skin lesions and may help to differentiate between early MF patches and benign mimics. Clinical data indicate that PCR will identify a dominant clonal T-cell receptor-γ rearrangement in 63% to 90% of skin biopsies that show definite histologic evidence of MF.^47,⁴⁸ Furthermore, PCR identifies a clonal T-cell population in 50% to 80% of histologically borderline biopsies obtained from patients who subsequently develop classic MF.^25,²⁶ In contrast, T-cell clonality occurs in only 6% to 24% of benign dermatoses that contain a lymphoid infiltrate.^47,⁴⁸ These observations suggest that identification of a clonal T-cell population should always be considered in light of the clinical and histologic context and may help to confirm a diagnosis of MF when it is already suspected on these grounds.

Transformation to Large Cell Histologic Variant

Transformation to a large cell variant occurs in up to 39% of patients initially diagnosed with MF.⁴⁹ Histologic diagnosis of transformation can be made when large cells (four times or more the size of a small lymphocyte) make up more than 25% of the lymphoid infiltrate or form microscopic nodules.⁵⁰ Transformation is associated with expression of CD30 in 30% of cases and expression of CD20 in 45% of cases. Transformed MF is typically aggressive, with clinical behavior similar to that of high-grade lymphoma.

Clinical Manifestations

According to the EORTC classification of cutaneous lymphomas, the term mycosis fungoides should be reserved for those CD4+ cutaneous lymphomas that are “characterized by the subsequent evolution of patches to more infiltrated plaques and eventually tumors.”² Over time, MF may spread to involve lymph nodes, blood, bone marrow, and visceral organs. Symptoms may vary by degree of involvement, but weight loss, night sweats, and fever are uncommon unless infection is present.

Premycotic Phase

Early MF typically begins with mildly erythematous, slightly scaling, annular or arcuate macules that classically involve sun-shielded areas (Fig. 77-3A). These lesions may wax and wane for years before histologic findings show definitive evidence of MF.

Figure 77-3 Spectrum of cutaneous lesions observed in mycosis fungoides. A, Typical early patch with erythema and mild scale. B, Typical plaque, with raised, palpable borders, central clearing, and overlying scale. C, Large tumor with necrosis and ulceration. D, Generalized erythroderma.

Reprinted with permission from Figure 1 in Smith BD, Wilson LD: Management of mycosis fungoides. Part 1. Diagnosis, staging, and prognosis. Oncology 17:1281-1288, 2003.

Patch Phase

In this stage, patches lose their predilection for sun-shielded areas and may become eczematous, hypopigmented, or hyperpigmented. The trunk, pelvis, and proximal extremities are most commonly involved. Histologic features consistent with MF may now be discernible.

Plaque Phase

If untreated, some patches will progress to form more generalized, deeply infiltrative, scaling plaques that often have well-demarcated, palpable borders and may exhibit central clearance and arcuate morphologic characteristics (see Fig. 77-3B). Associated findings include hyperkeratosis of the palms and soles and fissures.

Tumor Phase

Over 80% of tumors emerge in the setting of established patch-plaque MF (see Fig. 77-3C). The most common sites of tumor involvement include the face, digits, and perineum. Tumors frequently ulcerate and are prone to infection.

Tumeur d’Emblée

The term tumeur d’emblée refers to the rare patient who presents with tumors that arise in the absence of antecedent skin lesions; some cases may be more appropriately classified as CD30− cutaneous large T-cell lymphoma rather than MF.⁵¹ The clinical course may be more aggressive than that of classical MF.

Erythroderma

Erythroderma is defined as greater than 80% body surface involvement with confluent patches and/or plaques. It is associated with intense pruritus, hyperkeratosis of the palms and soles, skin atrophy, and lichenification (see Fig. 77-3D). Erythroderma may arise de novo or from progression of patch/plaque MF.

Lymph Nodes

Lymph node involvement is present in 15% of newly diagnosed patients and is associated with advanced cutaneous disease.⁵² Nodes are typically nontender and mobile and measure from 1 to 4 cm. Bulky adenopathy is uncommon.

Internal Organs

Visceral involvement is typically seen only in patients with advanced cutaneous disease, nodal disease, and blood involvement. The most common sites include the lungs, central nervous system, oral cavity, and oropharynx,52,53 although MF has been observed in other sites such as the breast, thyroid, and pancreas. Visceral involvement is often subclinical and does not routinely precipitate death. Bone marrow involvement at initial staging has been reported in 6% to 28% of patients and is also associated with advanced skin and nodal disease.^54,⁵⁵

Sézary Syndrome

Sézary syndrome (SS) is defined as erythroderma plus evidence of malignant circulating T cells that satisfy any of the five criteria ⁵⁶ listed in Table 77-3. For the purposes of these criteria, the Sézary cell is defined as “any atypical lymphocyte with [a] moderately to highly infolded or grooved nucleus”⁵⁶ (see Fig. 77-2F). Clinical findings may include edema and tumorous involvement of the face leading to leonine facies, severe fissures of the palms and soles, intense pruritus, and cutaneous pain.

TABLE 77-3 Proposed Hematologic Criteria for Diagnosis of Sézary Syndrome

Absolute Sézary cell count ≥1000 cells/µL.

CD4/CD8 ratio ≥10 due to an increase in CD3+¹ or CD4+¹ cells by flow cytometry.

Aberrant expression of pan-T-cell markers (CD2, CD3, CD4, CD5) by flow cytometry. Deficient CD7 expression on T cells (or expanded CD4+¹, CD7− cells ≥40%) is a tentative criterion.

Increased lymphocyte count with T-cell clone in blood identified by Southern blot or polymerase chain reaction.

A chromosomally abnormal T-cell clone.

Adapted from Vonderheid EC, Bernengo MG, Burg G, et al: Update on erythrodermic cutaneous T-cell lymphoma. Report of the International Society for Cutaneous Lymphomas. J Am Acad Dermatol 46:95-106, 2002.

Currently, no consensus has emerged regarding the pathologic link between MF and SS. However, some clinical observations help to elucidate the relationship between these two disorders. SS usually arises de novo without antecedent MF. Rarely, MF will evolve to an erythrodermic stage with concomitant hematologic findings that satisfy a diagnosis of SS; these cases should be considered “SS preceded by MF.” Most patients with a history of MF who develop erythroderma do not meet hematologic criteria for SS and are therefore considered to have erythrodermic MF.

Variants of Mycosis Fungoides

The progression from subtle patches to indurated plaques, cutaneous tumors, and erythroderma represents classical so-called Alibert-Bazin mycosis fungoides. According to the EORTC classification, variants such as bullous and hyperpigmented or hypopigmented MF manifest similar clinical behavior and should not be considered separately from classical MF.² Several variants sharing some clinical and pathologic features with MF have been described.

Folliculotropic Mycosis Fungoides

This entity presents with follicular papules, comedo-like lesions, milia-like lesions, patches, and plaques, all of which may produce alopecia.^1,³⁸ Although it typically involves the head and neck, any skin site may be affected.² Pathologically, atypical lymphocytes invade follicles and may deposit acid mucopolysaccharides in the pilosebaceous units. Folliculotropic MF tends to be refractory to topical treatments such as PUVA and nitrogen mustard, and risk of relapse after treatment with TSEBT appears to be higher when compared with that of patients with classical MF.⁵⁷ The 5-year disease-specific survival (DSS) is approximately 80%.³⁸

Pagetoid Reticulosis or Woringer-Kolopp Disease

Pagetoid reticulosis, also known as Woringer-Kolopp disease, typically presents as a slow-growing, hyperkeratotic or psoriasiform, localized patch or plaque involving a distal extremity. Pathologically, an abundant epidermotropic infiltrate composed of atypical large lymphocytes is noted, along with pagetoid spread of individual lymphocytes interspersed among keratinocytes. Benign-appearing small lymphocytes are found in the upper dermis.¹ The prognosis for localized pagetoid reticulosis is excellent with either surgery or radiotherapy, and disease-related deaths have not been reported.^2,³⁸ Ketron-Goodman type is a disseminated, more aggressive cutaneous lymphoproliferative disorder that histologically resembles localized pagetoid reticulosis.

Granulomatous Slack Skin

This rare variant presents with lax skin in the axillae, neck, breasts, and inguinal regions. Histologic features include epithelioid or giant cell dermal granulomas and associated destruction of elastin fibers.² Notably, Hodgkin’s disease has been associated with granulomatous slack skin in roughly one-third of reported cases.⁵⁸ Because of its rarity, optimal treatment has not been established.

Related Cutaneous T-Cell Lymphoproliferative Disorders

In addition to MF and its variants, several other distinct cutaneous T-cell lymphoproliferative disorders have been described. Their clinical and pathologic features and optimal therapy are briefly reviewed in the following sections.

Primary Cutaneous Anaplastic Large Cell Lymphoma or CD30+ Large Cell Cutaneous T-Cell Lymphoma

This entity typically presents as a red or flesh-colored nodule or tumor that frequently ulcerates. Histopathologic testing shows sheets of CD30+ large lymphocytes without epidermotropism. In contrast to systemic CD30+ anaplastic large cell lymphoma, overexpression of anaplastic lymphoma kinase (ALK) is not found in primary cutaneous CD30+ large cell lymphoma. Patients with localized disease are typically treated with radiotherapy alone, and the prognosis is excellent, with a 5-year DSS of approximately 90% to 95%.2,59,60

Lymphomatoid Papulosis

Lymphomatoid papulosis (LyP) presents with grouped erythematous or violaceous papules and/or nodules at different stages of development. Lesions typically resolve spontaneously within 2 to 8 weeks, but scarring is common.⁶¹ Three different histologic subtypes have been described, with types A and C consisting of malignant CD30+ T cells, often with an extensive inflammatory infiltrate, and type B simulating classical plaque-stage MF. For type C lesions, discrimination between LyP and CD30+ large cell lymphoma may be difficult on histologic grounds, and assessment of the clinical context may be required to ensure the proper diagnosis.² Although cytologically malignant, it should be emphasized that LyP is clinically benign, with a 5-year survival rate of 100%.^2,38,59 Neither aggressive chemotherapy nor radiotherapy are therefore indicated.⁶² Treatment options include PUVA or low-dose methotrexate, but neither is considered curative. In the long term, at least 15% to 20% of patients with LyP will develop a secondary malignant tumor, most commonly MF, CD30+ large cell lymphoma, or Hodgkin’s disease.^62,63 In patients undergoing TSEBT for MF, a history of LyP is associated with an increased risk of relapse.⁵⁷

Adult T-Cell Lymphoma or Leukemia

Adult T-cell lymphoma/leukemia (ATLL) develops in 2% to 4% of individuals infected with human T-cell lymphotrophic virus type I (HTLV-I), a retrovirus endemic in southern Japan and the Caribbean.1,61,64,65 Skin findings are present in up to 60% of patients with ATLL and strongly resemble those of MF, including plaques, tumors, and erythroderma. The characteristic immunophenotype is CD2+ CD3+ CD4+ CD5+ CD7− CD8− and the malignant cells strongly express the high-affinity interleukin-2 receptor (CD25, CD122, and CD132). Clinical features vary from an acute form presenting with B symptoms, hypercalcemia, metabolic bone disease, hepatosplenomegaly, generalized adenopathy, and leukemic infiltration to a smoldering or chronic form presenting with skin infiltration and little or no systemic involvement. Although patients often respond to conventional chemotherapy, long-term survival is rare. Newer promising agents include denileukin diftitox⁶⁶ and interferon-α in combination with zidovudine.⁶⁷

Subcutaneous Panniculitis-like T-Cell Lymphoma (α/β Type)

In prior years, subcutaneous panniculitis-like T-cell lymphoma was thought to run two markedly different clinical courses—an indolent course and an aggressive course. Recently, molecular studies have indicated that the indolent form of subcutaneous panniculitis-like T-cell lymphoma is caused by CD8+ T cells that express the α/β T-cell receptor. This entity typically presents with subcutaneous nodules or plaques involving the legs or trunk and has a 5-year DSS of approximately 82%.³⁸ In contrast, the more aggressive form of subcutaneous panniculitis-like T-cell lymphoma is caused by CD8− T cells that express the γ/δ T-cell receptor and is now classified as cutaneous γ/δ T-cell lymphoma in the updated WHO-EORTC system.³⁸ This entity is often fatal and complicated by hemophagocytic syndrome. Treatment for the α/β type is often observation, whereas treatment for the γ/δ type includes combination chemotherapy and possibly cyclosporine.^1,²

Prognosis and Staging

The seventh edition of the American Joint Committee on Cancer (AJCC) staging system for MF identifies the extent and character of skin lesions, extracutaneous disease, and leukemic transformation as the major predictors of poor prognosis ⁶⁸ (Tables 77-4 and 77-5). Shortcomings of this system include (1) difficulty in assigning T category for patients who fall on the border between T1 and T2; (2) failure to discriminate between the prognosis of patients with extensive patches as compared with extensive plaques; (3) similarity in prognosis between patients with tumors and erythroderma; (4) the questionable prognostic relevance of enlarged, pathologically uninvolved lymph nodes; and (5) the rarity of the N2 descriptor because biopsies of nonpalpable lymph nodes are seldom performed.^69,⁷⁰ To address these shortcomings, the EORTC, in conjunction with the International Society for Cutaneous Lymphoma, recently proposed a new staging system for MF, though it remains to be seen how this proposal will affect subsequent iterations of the AJCC staging system.⁷¹

TABLE 77-4 TNM(B) Classification for Mycosis Fungoides

T1	Patches and/or plaques involving <10% body surface area
T2	Patches and/or plaques involving ≥10% body surface area
T3	One or more cutaneous tumors
T4	Generalized erythroderma
N0	Lymph nodes clinically uninvolved
N1	Lymph nodes clinically enlarged but histologically uninvolved
N2	Lymph nodes clinically nonpalpable but histologically involved
N3	Lymph nodes clinically enlarged and histologically involved
M0	No visceral disease present
M1	Visceral disease present
B0	No circulating atypical cells (<1000 Sézary cells [CD4+¹ CD7–]/µL)
B1	Circulating atypical cells present (≥1000 Sézary cells [CD4+¹ CD7–]/µL)

Adapted from Edge SB, Byrd DR, Compton CC, et al: AJCC Cancer Staging Handbook, 7th ed. New York, Springer, 2010.

TABLE 77-5 Staging Classification for Mycosis Fungoides *

IA	T1N0M0
IB	T2N0M0
IIA	T1-2N1M0
IIB	T3N0-1M0
IIIA	T4N0M0
IIIB	T4N1M0
IVA	T1-4N2-3M0
IVB	T1-4N0-3M1

* The B descriptor is not considered in stage classification.

Adapted from Edge SB, Byrd DR, Compton CC, et al: AJCC Cancer Staging Handbook, 7th ed. New York, Springer, 2010.

Data from a cohort of 468 patients with newly diagnosed MF evaluated at Stanford University ⁵² are presented in Table 77-6. Over 60% of patients presented with patches and/or plaques without nodal or visceral disease. Such patients rarely developed disseminated disease, even at 20 years of follow-up. In contrast, 50% to 60% of patients presenting with tumors or erythroderma will develop extracutaneous disease.

TABLE 77-6 Initial Stage and Natural History of 468 Patients with Mycosis Fungoides Evaluated at Stanford University at the Time of Initial Diagnosis

Data from Stanford University showed that patients with limited patches and plaques (IA, T1N0M0) experience 10-year survival rates similar to those of a matched control population. In contrast, the median survival time for patients with extensive patches and plaques (T2), tumors (T3), and erythroderma (T4) was 11, 3.2, and 4.6 years, respectively.⁷² Patients with either pathologically documented lymph node involvement or visceral involvement experienced a median survival time of roughly 1 year.⁵²

A multi-institutional study showed that survival time is also related to the pattern and extent of lymph node involvement.⁷³ For example, partial or complete effacement of nodal architecture by malignant lymphocytes conferred a median survival time of 2.3 years. In contrast, lymph nodes with aggregates of atypical lymphocytes but preserved nodal architecture resulted in a median survival time of 6 years and lymph nodes with only dermatopathic changes or few atypical lymphocytes resulted in a median survival time of 9 years.

Other factors that may herald a poor prognosis include patient age of 60 years or more,⁷⁴ elevated LDH levels,⁷⁴ elevated soluble interleukin-2 receptor levels,³⁷ a low percentage of CD8+ tumor infiltrating lymphocytes,⁷⁵ extent of skin involvement in those with T3 disease,⁷⁶ T-cell clonality within the cutaneous infiltrate detected by PCR,^77,⁷⁸ an identical T-cell clone in the skin and peripheral blood,⁷⁹ and T-cell clonality in dermatopathic lymph nodes.⁸⁰

The AJCC also included a blood descriptor, B0 versus B1, to document the absence or presence of more than 1000 Sézary cells (CD4+, CD7− per µL). Among patients with erythroderma, the presence of B1 disease is a significant, independent predictor of survival time, doubling the risk of death.⁸¹

For those patients with transformation to a large cell variant of MF, the median survival time ranges from 19 months to 36 months.^49,⁵⁰ Factors predictive of poor survival after transformation include a short interval between the diagnosis of MF and transformation (<2 years) and the presence of IIB to IV disease.⁴⁹

Patient Evaluation

History

All patients should have a thorough history performed by the evaluating dermatologists, radiation oncologists, and any other consultants (Table 77-7). Careful attention should be given to the duration, change in appearance, and distribution of the eruption according to the patient. Inquiry related to a history of pruritus, pain, exfoliation, fissures, bullae, and perineal discomfort should be made. Previous diagnostic considerations, procedures, and therapies should be recorded in detail to establish temporal relationships. The clinical interview, along with the examination, are the two most important aspects of the workup because they can help exclude other diagnoses and establish whether the narrative and findings are consistent with a diagnosis of MF.

TABLE 77-7 Recommended Workup for Patients with Suspected Mycosis Fungoides

History and physical examination with attention to skin, lymph nodes, liver, and spleen

Biopsy of enlarged lymph nodes with attention to

• Histologic findings, both number of atypical lymphocytes and disruption of nodal architecture

• In dermatopathic nodes, consider PCR for T-cell clonality and immunophenotyping to rule out occult involvement

Evaluation of blood

• CBC with manual differential, liver function tests, and serum chemistries for all patients

• In those with suspected stage IIB to IV disease

• LDH, soluble interleukin-2 receptor

• Flow cytometry for CD2, CD3, CD4, CD5, CD7, CD8, CD20, CD45RO

• PCR for T-cell receptor gene rearrangement

Imaging

• Posteroanterior and lateral chest radiograph for stages IA to IB

• CT of chest, abdomen, and pelvis for suspected stage IIA to IV

• Consider PET/CT scanning for patients with suspected stage IIA to IV

CBC, complete blood count; CT, computed tomography; LDH, lactate dehydrogenase; PCR, polymerase chain reaction; PET, positron emission tomography.

Physical Examination

All patients should undergo a complete physical examination with special attention to the skin surface, lymph nodes, and abdomen. Specific notations should be made of the number, location, character, and distribution of cutaneous lesions. The percentage of cutaneous involvement should also be quantified. Other non-MF cutaneous findings, such as previous excision sites, pigmented lesions, or any findings consistent with other skin malignant tumors, should be noted. Adenopathy should be carefully documented with respect to location, size, consistency, mobility, and discomfort on palpation. Photographs of the skin for baseline documentation should be considered.

Diagnostic Testing

Pathology

A minimum of two biopsies of involved skin should be performed for hematoxylin and eosin evaluation. Material for immunophenotyping and T-cell receptor gene rearrangement should be obtained. Excisional biopsy of adenopathy is recommended, and tissue should be submitted for the same pathologic studies. Needle aspiration of nodes is not recommended.

Routine bone marrow examination is generally unnecessary but may be considered in patients with blood or other visceral involvement. Bone marrow involvement at initial staging has been reported in 6% to 28% of patients and is associated with advanced skin and nodal disease.^54,⁵⁵ There is no evidence to suggest that marrow involvement is an independent predictor of outcome, however.^55,74 Furthermore, although a clonal T-cell population within the marrow is identified in 75% of patients with identical skin and peripheral blood T-cell clones, this finding does not appear to alter the prognosis.⁵⁴

Hematology and Chemistry

A complete blood count, including differential and smear, in addition to serum chemistries (to include renal and hepatic function) should be obtained prior to initiation of therapy. For patients with IIB to IV disease, serum levels of soluble interleukin-2 receptor and lactate dehydrogenase (LDH) reflect the overall tumor burden and can be used to assess the response to therapy.

A thorough examination of the peripheral blood for evidence of malignant T cells is indicated for those with IIB to IV MF and may be considered for those with IA to IIA disease. Peripheral blood flow cytometry should be performed to assess expression of CD2, CD3, CD4, CD5, CD7, CD8, CD20, and CD45RO. Findings suggestive of blood involvement include an elevated CD4 : CD8 ratio (normal range, 0.5 to 3.5), or an expanded population of CD4+D7− or CD45RO+ lymphocytes.⁵⁶ If initially abnormal, findings on peripheral blood flow cytometric testing should be followed to assess the response to therapy. If flow cytometry is unremarkable, peripheral blood PCR for T-cell receptor gene rearrangement should be considered. PCR reveals a clonal T-cell population identical to that found in the cutaneous infiltrate in 40% of patients with erythroderma and 14% of patients with patches, plaques, or tumors.⁴⁷

Diagnostic Imaging

Chest radiography should be completed for all patients via posteroanterior and lateral views. CT of the chest, abdomen, and pelvis may be obtained for patients with tumors, erythroderma, or nodal involvement but is not necessary for those with IA or IB disease given the low yield.⁸² FDG-PET (fluorodeoxyglucose–positron emission tomography) may also be beneficial for patients with advanced skin disease; for example, Stanford University reported that CT alone identified pathologic adenopathy by size criteria in only 5 of 13 patients, whereas the addition of PET identified hypermetabolic adenopathy in 13 of 13 patients. The standardized uptake value (SUV) of the FDG correlated with the extent of nodal involvement, with the highest SUVs noted in patients with complete nodal effacement and large cell transformation.⁸³ To confirm nodal involvement and document the presence or absence of transformation, excisional biopsy should be strongly considered for selected abnormal lymph nodes that are radiographically identified.

Primary Therapy

Because MF is a disease of cutaneous (i.e., skin-homing) lymphocytes, therapy is quite distinct from that of nodal lymphomas. For patients with disease limited to the skin, topical therapy alone produces high rates of remission and even cure. Topical therapies include mechlorethamine (nitrogen mustard), carmustine (BCNU), steroids, bexarotene gel, psoralen plus ultraviolet A (PUVA), ultraviolet B (UVB), and either localized or total skin electron radiotherapy. For patients with localized unilesional MF, topical therapies alone produce long-term DFS in excess of 85%.¹⁹^–²¹ For those with more than one patch or plaque but with less than 10% body surface area involvement (T1N0 and T1N1), topical therapies alone produce long-term DFS ranging from 30% to 50%. For patients with more extensive patches or plaques, topical therapies may produce remission, but long-term cure is unlikely. Such patients should receive intensive topical therapy to induce a complete remission followed by less intensive adjuvant topical therapy to sustain a remission.^84,⁸⁵

Those with tumors or erythroderma experience severe cutaneous symptoms and are at high risk for extracutaneous dissemination. Of all the topical therapies, total skin electron beam therapy (TSEBT) is associated with the highest rates of complete response for this patient subgroup.⁸⁵^–⁸⁷ Therefore it is recommended that TSEBT be administered to induce cutaneous remission and that adjuvant systemic and/or topical therapy be administered to sustain remission. Systemic therapies include interferon, retinoids, oral bexarotene, denileukin diftitox, vorinostat, extracorporeal photochemotherapy (photopheresis), and cytotoxic chemotherapy.

In the following section, we first discuss all nonradiation topical therapies and then discuss the various systemic therapies. Finally, we discuss the clinical data for radiation and compare this to clinical data for various topical and systemic therapies in order to develop evidence-based guidelines for the treatment of MF.

Skin-Directed Therapy

Mechlorethamine

Topical mechlorethamine hydrochloride (HN2), also known as nitrogen mustard, is an alkylating agent with proven activity in the treatment of MF patches and plaques. Mechlorethamine is typically applied daily and continued for at least 6 months after a complete response.⁶⁹ Cutaneous intolerance, manifested by erythema and pruritus, occurs in roughly 50% of patients treated with aqueous HN2,⁸⁸ but is reduced to less than 10% in patients treated with HN2 dissolved in ointment such as Aquaphor.⁶⁹ Other cutaneous side effects of HN2 may include xerosis, hyperpigmentation, and, rarely, bullous reactions, urticaria, and Stevens-Johnson syndrome.⁸⁸ Bone marrow suppression is not observed because of minimal systemic absorption. The carcinogenicity of HN2 remains debated because one series reported no increased risk of secondary skin cancers in patients treated with HN2 monotherapy,⁶⁹ but another series reported an eightfold increase in the risk of nonmelanoma skin cancers attributable to HN2 monotherapy.⁸⁹ HN2 may also potentiate the carcinogenicity of other topical therapies such as total skin irradiation or PUVA.⁹⁰

Carmustine

Topical carmustine (BCNU) is another alkylating agent with activity in MF. Because of systemic absorption that may produce bone marrow suppression, the drug should be applied to no more than 10% of the body surface area, the duration of treatment should be limited to 4 months, and complete blood counts should be monitored. Cutaneous hypersensitivity is uncommon (7% in one series), but chronic skin telangiectasias and hyperpigmentation may occur.⁹¹

Topical Steroids

High-potency topical glucocorticoids are an important component in the treatment of MF due to their ability to alleviate cutaneous symptoms and induce lesion regression. Typically, steroids are applied to active lesions only because widespread application can induce reversible depression of serum cortisol in 10% to 15% of patients.⁹² Persistent application can lead to skin atrophy.

Topical Rexinoids

Bexarotene belongs to a new class of agents called rexinoids that bind to the retinoid X receptor, resulting in transcription of various genes that control cellular differentiation and proliferation.⁹³ Topical bexarotene gel was recently approved after phase I and II trials demonstrated a 44% to 54% response rate in refractory cutaneous MF.^94,95 Most patients will develop an irritant dermatitis and thus require close observation and dose titration. Due to its irritant effects, bexarotene gel is not indicated for patients with more than 15% body surface area involvement. As with systemic retinoids, bexarotene in both its topical and systemic forms should be avoided in pregnant women due to possible teratogenicity.

Psoralen Plus Ultraviolet A (PUVA) and Ultraviolet B

Ultraviolet light used in the treatment of MF includes ultraviolet B (wavelength, 320 to 290 nm), narrow-band ultraviolet B (wavelength, 311 nm), or psoralen plus ultraviolet A (PUVA, wavelength, 400 to 320 nm). Because ultraviolet B has limited penetration, its efficacy is limited to thin patches. PUVA penetrates more deeply and will effectively treat some plaque lesions. PUVA requires ingestion of a photochemotherapeutic agent, 8-methoxypsoralen, prior to ultraviolet A exposure. Ultraviolet A activates 8-methoxypsoralen, resulting in DNA cross-linking and apoptotic cell death.⁹⁶

Both PUVA and ultraviolet B are initially administered 2 to 3 days a week as the light dose is gradually increased. Once a complete response has been achieved, treatment frequency may be gradually reduced to once every 2 to 4 weeks. Ultraviolet treatments may be continued on a maintenance basis for several years, provided the patient remains in complete response with treatment administered once every 4 to 8 weeks.

Acute side effects of PUVA and ultraviolet B include skin erythema that may be painful, hyperpigmentation, xerosis, pruritus, and blistering. Eye goggles are used to decrease the risk of cataract formation. One side effect unique to PUVA is nausea and vomiting after ingestion of 8-methoxypsoralen (8-MOP). This may be avoided by substituting 8-MOP for either a topical “psoralen bath” or 5-methoxypsoralen, a nonemetogenic analog that is currently available in Europe.⁹⁷ Long-term toxicity includes photoaging and increased risk of melanoma and nonmelanoma skin cancers.^98,⁹⁹

Systemic Therapy

Interferon

Interferon-α-2a (IFN-α) is an effective agent, particularly for patch and plaque disease, likely due to a direct antitumor effect and/or immunomodulation.¹⁰⁰ Interferon-α has been used alone^100,101 or in combination with retinoids, PUVA,^102,103,104 and extracorporeal photopheresis.¹⁰⁵ To date, no randomized trials have compared interferon-α plus another therapy with interferon-α alone, although one prospective trial did show a benefit from combining interferon-α with PUVA as compared with interferon-α plus a retinoid.¹⁰⁶ Toxicity may include flulike symptoms, psychiatric disturbances including depression and confusion, elevated transaminases, leukopenia, thrombocytopenia, proteinuria, and myelopathy.^103,104 Despite these side effects, in a recent phase II trial of interferon-α and PUVA, only 8% of patients withdrew as a result of toxicity.¹⁰⁴

Retinoids

Oral retinoids such as isotretinoin and acitretin influence cellular differentiation and may be particularly beneficial in the treatment of folliculotropic MF. Retinoids can be safely combined with other therapies such as PUVA,¹⁰⁷ interferon-α, and TSEBT.¹⁰⁸ Side effects include photosensitivity, xerosis, myalgias, arthralgias, headaches, impaired night vision, corneal opacities, teratogenicity, elevated transaminases, hyperlipidemia, and pancreatitis.⁷²

Rexinoids

Oral bexarotene has been approved by the U.S. Food and Drug Administration (FDA) for use in all stages of treatment-refractory MF.^109,¹¹⁰ Preliminary data suggest that bexarotene can be combined safely with other therapies, including PUVA, extracorporeal photopheresis, interferon-alpha, and HN2.

Hypertriglyceridemia, the most common adverse event, occurs in 80% of treated patients and may result in reversible pancreatitis if triglyceride levels exceed 800 mg/dL.⁹³ Therefore atorvastatin or fenofibrate should be initiated if triglyceride levels exceed 350 mg/dL. Gemfibrozil increases serum bexarotene concentrations, resulting in a paradoxic elevation of triglycerides, and should therefore be avoided.⁹³ Another side effect, central hypothyroidism, affects roughly 75% of patients but responds well to levothyroxine and resolves when treatment is discontinued.⁹³ Patients taking bexarotene therefore require monitoring of estimated free thyroxine because levels of thyroid-stimulating hormone (TSH) will always be low. Other side effects include self-limited headaches, mild neutropenia, mild transaminase elevations, skin peeling, and pruritus. In the phase II to III trial of oral bexarotene for advanced, refractory CTCL, only 10% of patients receiving the optimal dose withdrew as a result of an adverse event.¹⁰⁹

Denileukin Diftitox

Denileukin diftitox is a recombinant fusion protein that contains portions of interleukin-2 and diphtheria toxin and has proven activity in refractory MF, stages IB to IVA.¹¹¹ It selectively targets T cells that express the high-affinity interleukin-2 receptor (a complex of CD25, CD122, and CD132), resulting in endocytosis of diphtheria toxin, inhibition of protein synthesis, and cell death. In the pivotal phase III clinical trial leading to its approval, only patients with neoplastic T cells expressing the high-affinity interleukin-2 receptor received denileukin diftitox, and controversy currently exists as to whether or not patients without significant expression of the high-affinity interleukin-2 receptor benefit from this therapy.¹¹² Denileukin diftitox is typically administered by a 30-minute venous infusion given on five consecutive days and repeated every 3 weeks for up to eight cycles.

Toxicities were commonly encountered in the phase III clinical trial of denileukin diftitox. Acute hypersensitivity reactions, including dyspnea, back pain, hypotension, and chest pain, occurred in 60% of patients. Furthermore, a vascular leak syndrome characterized by hypotension, hypoalbuminemia, and edema was encountered in 25% of patients. Other toxicities may include constitutional symptoms, thrombotic events, infections, transaminase elevations, renal impairment, and lymphopenia. In total, 21% of the patients in this trial withdrew due to adverse events.¹¹¹ Subsequent investigations have suggested that pretreatment with corticosteroids substantially reduces the risk of acute toxicity.¹¹³

Vorinostat

Vorinostat is a histone deacetylase inhibitor (HDAC) approved for use in patients with MF who have progressive, persistent, or recurrent disease following treatment with at least two systemic therapies. HDAC inhibitors increase accumulation of acetylated histones, resulting in decreased availability of nuclear DNA to bind to transcription factors. The decrease in transcription decreases intracellular protein levels, ultimately producing cell cycle arrest and apoptosis. In preclinical studies of MF cell lines, vorinostat has been shown to induce apoptosis and downregulate stat6.¹¹⁴ Phase II clinical studies have reported clinical response rates of 25% to 30% in pretreated patients, with a modest median response duration of 15 to 26 weeks.^115,¹¹⁶ Common toxicities include diarrhea, nausea, fatigue, and anorexia.

Extracorporeal Photochemotherapy

Extracorporeal photochemotherapy (EP), also known as photopheresis, is a novel immune therapy that has shown activity in the treatment of erythrodermic MF.96,117 Peripheral blood leukocytes are harvested by leukapheresis, mixed with 8-methoxypsoralen, exposed to 2 J/cm² of ultraviolet A, and then reinfused into the patient. This results in DNA cross-linking and gradual apoptotic death of the circulating MF cells that were exposed to psoralen plus UVA. For unclear reasons, monocytes are resistant to the apoptotic effects of EP but instead are stimulated to become immature antigen-presenting dendritic cells due to the physical process of leukapheresis. Therefore, once treated leukocytes are reinfused into the bloodstream, activated dendritic cells may phagocytize remnants of the apoptotic MF cells, present their antigens on major histocompatibility class I, and stimulate expansion of antitumor CD8+ T cells.¹¹⁸

EP is typically administered on two consecutive days every 4 weeks, although the frequency may be increased in patients with extensive disease.¹¹⁹ For patients who achieve a complete response, therapy should be maintained for roughly 6 months and then gradually tapered. In general, EP is well tolerated, although hypotension, arrhythmias, and heart failure may occur due to fluid shifts. Patients with a history of cardiac disease therefore require close monitoring.¹¹⁷ EP has been safely combined with interferon-α and TSEBT.^105,^120,¹²¹

Chemotherapy

Systemic chemotherapy is typically reserved for refractory cutaneous disease, visceral disease, or large cell transformation. Chemotherapy is not typically used in the initial management of patients with MF because a randomized phase III trial comparing concurrent TSEBT and systemic chemotherapy to sequential topical therapy failed to show an improvement in DFS or OS.¹²²

For patients who may benefit from chemotherapy, two main strategies exist. The first relies on oral agents such as methotrexate,¹²³ etoposide, or chlorambucil. This strategy avoids the need for central venous lines, which are associated with a high risk of infection due to frequent bacteremia caused by open skin lesions. The second strategy relies on intravenously administered chemotherapy. One exciting new prospect is pegylated liposomal doxorubicin, an agent that tends to remain intravascular but will extravasate into the inflamed lesional skin of MF. A pilot study conducted on 34 patients with CTCL reported a complete response in 15, a partial response in 15, and 6 severe adverse events.¹²⁴ Another notable agent is the nucleoside analog gemcitabine, which has been shown to induce complete response rates of approximately 10% and overall response rates of 70% without an increased risk of infection.¹²⁵ Attention has also focused on purine analogs such as fludarabine, 2′-deoxycoformycin, and 2-chlorodeoxyadenosine. Clinical trials of these agents, however, have produced response rates ranging from only 28% to 51% and have documented substantial toxicity, including myelosuppression, infection, and pulmonary dysfunction.¹²⁶^–¹²⁹ Other agents with activity in MF include the lipophilic antifolate trimetrexate¹³⁰; 5-fluorouracil; cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP)¹³¹; cisplatin; etoposide; bleomycin; and vinblastine.¹²⁵

High-Dose Chemotherapy with Autologous or Allogeneic Bone Marrow Transplant

Limited experience has been reported using high-dose chemotherapy with either autologous or allogeneic transplant in the treatment of advanced refractory MF. Autologous stem cell rescue has resulted in several complete responses; however, relapses within 1 year are the rule rather than the exception.¹³²^–¹³⁴ In contrast, myeloablative chemotherapy with allogeneic bone marrow transplant may result in prolonged DFS times. One series of three patients treated with cytoreductive chemotherapy and total body irradiation (TBI) found that two of the three patients remained disease free at 4.5 years and 15 months after transplant. The third patient had recurrent disease after allogeneic transplant but developed a second complete response upon withdrawal of prophylactic cyclosporine, suggesting a graft-versus-tumor effect.^135,¹³⁶ In another series, six of eight patients with heavily pretreated, refractory MF who underwent allogeneic transplant with either TBI or non-TBI preparative regimens were alive without evidence of lymphoma at a median of 56 months following transplant, but two deaths secondary to toxicity were reported.¹³⁷ Nonmyeloablative transplants have also been reported, with one series reporting clearance of clonal T cells and durable complete remissions in three out of three patients, although one died from infectious complications.¹³⁸ Collectively, these experiences strongly support the use of allogeneic transplant for younger patients with treatment-refractory advanced MF.

Radiotherapy

Radiation therapy is the most effective single modality in the treatment of MF ⁸ and plays an important role in the treatment of localized or disseminated cutaneous disease and in the palliation of nodal and visceral metastases.

Dose

A dose response has been reported both for radiation of single MF lesions and for TSEBT. For example, in a study of 110 lesions from 14 patients with at least 1 year of follow-up, Cotter and colleagues ¹³⁹ reported an infield recurrence rate of 42% for those treated to a total dose of 10 Gy or less, 32% for 10.01 to 20 Gy, 21% for 20.01 to 30 Gy, and 0% for more than 30 Gy. In a study of 30 lesions from patients with stage IA MF treated with localized radiotherapy, Wilson and associates²⁰ reported local failure in 20% (4 of 20 patients) treated with 20 Gy or less, compared with 0% (0 of 10 patients) treated with more than 20 Gy.²⁰ Kim and coauthors¹⁴⁰ treated different lesions from the same patient with various total doses and fraction sizes, finding that division of the total dose into two fractions separated by 1 or 7 days did not alter local control rates. This suggests that the malignant cells of MF have minimal ability to execute sublethal damage repair and provides justification for using low daily fraction sizes to minimize normal tissue toxicity without sacrificing local control. Collectively, these experiences demonstrate that curative treatment of localized MF requires total doses of at least 20 Gy and perhaps up to 30 Gy. To spare normal tissue toxicity and to ensure that patients can receive TSEBT in the future, a fraction size of 1.2 to 2 Gy is recommended.

A dose-response relationship has also been demonstrated with TSEBT. A meta-analysis of published results reported higher rates of complete remission when using higher prescribed doses and more penetrating electrons.⁸⁶ Historically, the Stanford University experience with 176 patients undergoing TSEBT from 1958 to 1975 revealed that complete response rates increased with total dose: 18% for 8 to 9.9 Gy, 55% for 10 to 19.9 Gy, 66% for 20 to 24.9 Gy, 75% for 25 to 29.9 Gy, and 94% for 30 to 36 Gy.¹⁴¹ Similarly, patients treated with higher doses experienced improved OS, regardless of T category. The experience at Hamilton Regional Cancer Center between 1977 and 1992 further supported the importance of high-dose TSEBT. From 1977 to 1980, 25 consecutive patients received 30 Gy of TSEBT. From 1980 to 1992, 121 consecutive patients received 35 Gy of TSEBT.¹⁴² Treatment with high-dose TSEBT was an independent predictor of response, with a complete response rate of 64% for the 30-Gy group and 85% for the 35-Gy group. Given these data, the European Organization for the Research and Treatment of Cancer (EORTC) consensus statement recommended that modern TSEBT deliver a total dose of 31 to 36 Gy to the skin surface in order to produce a dose of 26 Gy at a depth of 4 mm in truncal skin along the central axis.¹⁴³

Clinical Results of Limited Superficial Radiotherapy

Approximately 5% of patients with stage IA disease present with a single skin lesion or with two or three lesions in close proximity, such that all clinically apparent disease can be encompassed by either one field or several abutting fields.²⁰ Radiotherapy is the treatment of choice in this situation, based on results from three institutions that have reported long-term DFS in excess of 85%.^20,^21,¹⁴⁴ Wilson and colleagues²⁰ published a series of 21 patients with minimal IA disease managed with local, superficial radiotherapy. Ten were treated with 100 to 280 kV and 11 with 4- to 12-MeV electrons with appropriate bolus. The median dose was 20 Gy, and 17 of 21 patients received 20 Gy or more. With a median follow-up of 36 months, the rate of complete clinical remission was 97% and the long-term DFS was 91% among patients receiving 20 Gy or more. Similarly, Micaily and associates²¹ reported an 86% 10-year DFS of in 18 patients with unilesional MF treated with localized radiotherapy to a median dose of 30.6 Gy. In summary, radiotherapy for unilesional IA disease is an excellent first-line therapy given minor acute and chronic toxicities and excellent long-term results. Because very few patients with two to four lesions have been reported in the literature, optimal treatment for this subgroup remains unclear. Although some advocate limited superficial radiotherapy, others advocate a strategy such as TSEBT or PUVA in order to treat all skin surfaces.

Clinical Results of Total Skin Electron Beam Therapy

Limited Patches and/or Plaques: T1N0M0 (IA) and T1N1M0 (IIA) Disease

Initial treatment options for patients with patches and/or plaques involving less than 10% of the body surface area include topical mechlorethamine, topical carmustine, phototherapy, bexarotene gel, and TSEBT. Currently, no randomized data exist to support selection of a radiation-based strategy over another topical strategy for this patient population. Although single-institution experiences suggest that patients managed with TSEBT may experience superior complete response rates and relapse-free survival rates, there is little strong evidence to suggest that this will translate into improved OS.

For example, patients with T1 disease treated with modern TSEBT experience a complete response rate of at least 90%^86,¹⁴⁵ (Table 77-8), compared with a complete response rate of 65% to 70% for topical mechlorethamine.^69,⁸⁹ Furthermore, the 10-year relapse-free survival rate is roughly 50% for TSEBT¹⁴⁶ compared with 34% for mechlorethamine.⁶⁹ In this group of patients, however, whose long-term survival rate is similar to that of healthy controls, initial treatment with TSEBT has not been associated with an OS advantage.¹⁴⁵ The administration of TSEBT for patients with IA MF remains controversial, and although some reserve TSEBT for those refractory to standard therapies, others recommend TSEBT as first-line therapy for even paucilesional IA disease.

TABLE 77-8 Newly Diagnosed Patch Plaque Mycosis Fungoides Managed at Hamilton, Ontario*

Other first-line therapies for T1 MF include phototherapy and carmustine. For patients with limited patches, ultraviolet B therapy results in a complete response rate of roughly 80%, with a median response duration of 2 years.147,148 PUVA is appropriate for both patches and plaques, producing complete response rates of at least 80%. Although relapse is common, most patients will respond to additional PUVA.^149,¹⁵⁰ Similar to mechlorethamine, topical BCNU produced a complete response rate of 86%, partial response rate of 12%, and 5-year relapse-free survival rate of 35%.⁹¹

For T1 disease that recurs after mechlorethamine, BCNU, ultraviolet B, or PUVA, retreatment with these agents may result in another response. Other options for recurrent or refractory T1 disease include TSEBT,¹⁴³ class I steroids (complete response, 63%; partial response, 31%),⁹² interferon-α monotherapy (complete response, 50%; partial response, 35%),¹⁰⁰ topical bexarotene (complete response, 21%; partial response, 42%),⁹⁴ and oral bexarotene (complete response, 7%; partial response, 47%).¹¹⁰

For patients whose disease recurs after initial treatment with TSEBT, relapse is typically limited to less than 5% of the skin surface. Salvage treatments including spot radiation, topical steroids, mechlorethamine, or PUVA have all resulted in prolonged second remissions. For example, data from Hamilton showed that 70% of patients were free of disease 15 years after first-line TSEBT and second-line salvage treatments as needed.¹⁴⁶

Extensive Patches and/or Plaques: T2N0M0 (IB) and T2N1M0 (IIA) Disease

For patients in this group with either severe cutaneous symptoms or deeply infiltrative plaques, we favor TSEBT as initial treatment due to superior response rates that produce rapid palliation. For example, complete response rates for T2 disease range from 76% to 90% (see Table 77-8), compared with a complete response rate of 34% for topical mechlorethamine.^69,^85,¹⁴³ Adjuvant therapy following TSEBT is crucial for patients with T2 disease because the 10-year relapse-free survival rate is only 10% in those treated with TSEBT alone.¹⁴⁶ Retrospective data from Yale University found that patients receiving adjuvant PUVA experienced a 5-year DFS of 85% versus 50% for those not receiving adjuvant PUVA.⁸⁴ A prospective pilot study at Hamilton documented a similar benefit (see Table 77-8). Retrospective data from Stanford University showed that patients treated with TSEBT and adjuvant topical mechlorethamine experienced a 10-year relapse-free survival rate of roughly 40% compared with 10% for TSEBT alone.⁸⁵

Patients with T2 disease who are relatively asymptomatic can receive treatment similar to that recommended for T1 disease. Again, topical mechlorethamine is a reasonable choice for these patients, resulting in a complete response rate of 34%, partial response rate of 38%, and 10-year relapse-free survival rate of 20%.⁶⁹ Similarly, BCNU produced a complete response rate of 47%, partial response rate of 37%, and 5-year relapse-free survival rate of 10%.⁹¹ Topical corticosteroids may also be helpful for patients with extensive patches, with data suggesting a complete response rate of 25% and a partial response rate of 57%.⁹²

As an alternative to topical chemotherapy, topical phototherapy may be considered for extensive patch/plaque disease. Ultraviolet B should be reserved for extensive patch disease only, due to its superficial penetration. In contrast, PUVA may be considered for patients with patches or thin plaques and has resulted in complete response rates of 60% to 100%.^92,^149,¹⁵⁰ Phase II trials suggest promising improvements in both the complete response rate and the duration of remission when adding interferon-α to PUVA for the treatment of T2 disease.^103,104,106

Cutaneous Tumors: T3N0-1M0 (IIB) Disease

Radiotherapy is an important treatment modality for cutaneous tumors due to its ability to treat the full thickness of deeply infiltrative lesions. For the rare patient with asymptomatic tumors involving less than 10% of the skin, either topical HN2 with local radiotherapy or TSEBT are reasonable first-line treatment options, with both producing a similar 5-year OS of roughly 50%.⁸⁵ Most patients with T3 disease present with extensive, symptomatic tumors, however. Such patients often benefit from TSEBT as a first-line, palliative treatment, due to a superior complete response rate as compared with topical HN2 plus localized radiotherapy (44% to 54% vs. 8%).^85,⁸⁶

Adjuvant therapy should be strongly considered for patients who achieve a complete response to TSEBT. Retrospective data suggest that adjuvant topical mechlorethamine may improve the duration of response, resulting in a 5-year relapse-free survival rate of 55% compared with 30% with TSEBT alone.⁸⁵ Adjuvant photopheresis is another reasonable option, with retrospective data reporting a 5-year OS of 100% for patients receiving this modality compared with 50% in patients who did not receive adjuvant therapy.¹²⁰ Other adjuvant therapies worthy of consideration include interferon-α, bexarotene, and denileukin diftitox.¹⁵¹

Erythroderma: T4N0-1M0 (III) Disease

TSEBT is an appropriate initial therapy for erythrodermic MF due to its ability to produce a rapid and sustained response, thereby ameliorating the severe cutaneous symptoms experienced by such patients. Further, limited early data suggest that TSEBT can result in a substantial reduction in the number of malignant cells circulating in the peripheral blood, potentially altering the natural history of the disease.¹⁵² Retrospective data indicate that TSEBT monotherapy produced a 100% complete response rate and 5-year PFS of 69% for patients with T4N0M0B0 MF disease.⁸⁷ When including those with blood or visceral involvement, however, the complete response rate dropped to 74% and only 36% remained progression free at 5 years. Patients with erythroderma and blood or visceral involvement may benefit from adjuvant photopheresis, as retrospective data suggest that such treatment improves the 2-year cause-specific survival rate from 69% without EP to 100% with EP.¹²¹

For patients without access to TSEBT, several other topical treatment options exist. For example, a series of 10 patients treated with PUVA reported a complete response rate of 70% and a median PFS of 5 months.¹⁵³ Although not studied in a randomized setting, prospective phase II experiences suggested that addition of interferon-α to PUVA may improve the duration of response.^102,103,104 Another option for T4 disease is photopheresis monotherapy, following which 80% of patients will experience at least some cutaneous improvement.⁹⁶ Patients with a normal peripheral blood CD4/CD8 ratio appear more likely to respond.¹¹⁹

Toxicity of Total Skin Electron Beam Therapy

TSEBT is generally well tolerated, and toxicity is minimized by using low daily fraction sizes ¹⁵⁴ and a shielding regimen that reduces the dose to eyes, ears, lips, hands, and feet. Common acute toxicities from TSEBT include pruritus, dry desquamation, erythema, alopecia, xerosis, bullae of the feet, edema of the hands and feet, hypohidrosis (diminished perspiration),¹⁵⁵ and loss of fingernails and toenails.^146,156 Rare acute side effects include gynecomastia in men, mild epistaxis, and mild parotiditis.¹⁴⁶ Because of the superficial penetration of electrons, patients do not experience gastrointestinal or hematologic toxicities. In general, TSEBT does not cause serious long-term complications,¹⁵⁷ although permanent nail dystrophy, xerosis, telangiectasias, partial scalp alopecia, and fingertip dysesthesias have been described.¹⁴³ Secondary cutaneous malignant diseases, including squamous cell carcinoma, basal cell carcinoma, and malignant melanoma, have been observed in patients treated with TSEBT, particularly in those exposed to multiple therapies that are themselves known to be mutagenic, such as PUVA and mechlorethamine.^90,158

Follow-up

Follow-up after definitive treatment for MF is best carried out by a multidisciplinary team with expertise in CTCL. Because most patients managed with TSEBT now receive adjuvant treatment, they remain in close contact with their dermatologists every 1 to 3 months for many years. Biopsy of clinically borderline or suspicious lesions is important to document relapse and to rule out transformation to a large cell variant. In addition, close surveillance is required to detect treatment-related skin cancers and ensure that appropriate therapy is initiated promptly. Patients with an elevated CD4 : CD8 ratio, LDH, or soluble interleukin-2 receptor level at initial diagnosis should have these values followed closely to determine response to therapy.

Locally Advanced Disease and Palliation

Young patients with locally advanced disease and a good performance status should be treated with curative intent with an allogeneic transplant when possible. However, for the majority of patients whose age or performance status precludes allogeneic transplant, treatment options include novel biologic agents, cytotoxic chemotherapy, and radiation therapy. For example, oral bexarotene has shown efficacy in patients with relapsed or refractory MF, producing an overall response rate of 57% for IIB, 32% for III, 44% for IVA, and 40% for IVB.¹⁰⁹ Another option, denileukin diftitox, resulted in an overall response rate of 30% for patients with treatment-refractory IB to IVA disease.¹¹¹ Systemic chemotherapy should also be considered and may result in complete response rates of 20% to 60%.¹³¹

Radiotherapy plays an important role in the palliation of both cutaneous and extracutaneous disease. For those patients with extensive skin disease recurrent after TSEBT, a second course of TSEBT may produce substantial palliation with acceptable toxicity. At Yale University, 14 patients have received two courses of TSEBT and 5 have received three courses of TSEBT.¹⁵⁹ The median dose was 36 Gy for the first course, 18 Gy for the second, and 12 Gy for the third. After the second course, 86% achieved a complete response, with a median disease-free interval of 11.5 months. In a similar experience from Stanford University, 15 patients received a second course of TSEBT to a median dose of 23 Gy.¹⁶⁰ The complete response rate was 40% and the partial response rate was 60%. Toxicity was limited to xerosis, telangiectasias, pigment changes, and alopecia. Criteria for retreatment include a complete response to the initial course, an extended disease-free interval after the initial course, diffuse cutaneous involvement at relapse, and failure of other modalities.

For patients with localized, symptomatic skin disease, radiotherapy options include higher doses on the order of 20 to 30 Gy, or lower doses on the order of 8 Gy in two fractions. Recent evidence suggests that the lower-dose approach produces complete response rates in excess of 90%, and that retreatment with higher doses of irradiation is safe and effective for those patients whose disease recurs following low-dose irradiation.¹⁶¹ Finally, patients with symptomatic nodal or visceral disease often benefit from a course of palliative megavoltage radiotherapy to a total dose of 20 to 30 Gy delivered in 2- to 3-Gy fractions.

Irradiation Techniques

Historical Development

Localized irradiation was first used to treat MF in 1902, shortly after the discovery of x-rays.¹⁶² In 1939, Summerville reported treatment of extensive cutaneous MF with an “x-ray bath” of kilovoltage photons delivered with two large fields.¹⁶³ A total air exposure of 900 R delivered in daily fractions of 10 R produced a complete response. In 1945, Levin and Behrman¹⁶⁴ proposed that the total air exposure at the skin surface should be 600 to 800 R for patches or plaques and 1000 to 1600 R for tumors.¹⁶⁴

The development of TSEBT began in the early 1950s. Trump and colleagues ¹⁶⁵ at the Massachusetts Institute of Technology used a Van de Graaff generator to produce a vertically oriented, stationary beam of 2.5-MeV electrons incident on a motorized couch. By placing the patient in the prone, supine, and lateral decubitus positions and translating the couch through the electron beam, all skin surfaces could receive a meaningful dose. This treatment approach produced complete responses in two patients with extensive MF. In 1960, Stanford University reported the first method for linear accelerator-based TSEBT.¹⁶⁶ With a patient standing 10 feet from the end of the accelerator, two fields were treated, one directed above the patient’s head and a second directed below the patient’s feet. This approach is known as the dual-field technique. In addition, patients were treated in two positions with respect to the accelerator: anteroposterior and posteroanterior. Over time, it became clear that increasing the number of treatment positions improved dose homogeneity in the lateral dimension.^167,¹⁶⁸ Ultimately, the Stanford group adopted a six-treatment position technique in which patients stand in six different orientations with respect to the accelerator: anteroposterior, posteroanterior, right and left anterior oblique, and right and left posterior oblique.¹⁶⁹ Although commonly called the six-field technique, it should be remembered that two fields are actually treated for each of the six treatment positions.

The first clinical results of TSEBT were published in 1962 by a group from St. John’s Hospital for Diseases of the Skin in London.¹⁷⁰ Using the dual-field, four-treatment position technique developed at Stanford, all five patients treated for MF experienced “very good” responses at total doses ranging from 12 to 18 Gy. Four of five patients relapsed within 8 months, however, and disease was difficult to control in the axillae and perineum due to underdosing. In 1971, Stanford University reported 107 patients with MF who received TSEBT between 1957 and 1968.¹⁶⁹ Using a variety of doses, they produced a 52% complete response rate. In addition, among patients with localized patches or plaques, 30% experienced long-term DFS, providing the first evidence that MF could actually be cured.

Target Volumes

The target volume for patients with patch/plaque disease should include the epidermis and dermis.⁸⁶ The thickness of the epidermis varies from 0.05 to 0.5 mm and is greatest in the distal extremities. The thickness of the dermis varies from 1 to 4 mm and is greatest in the hands and feet. Therefore, the thickness of the skin varies from a minimum of about 2 mm on the trunk to a maximum of about 4.5 mm at the hands and soles of the feet.⁸⁶ As a result, the EORTC TSEBT consensus statement recommends that the 80% isodose line should be 4 mm deep or more to the skin surface to ensure that the epidermis and dermis fall within the high-dose region. Due to their thickness, the deep margin of cutaneous tumors is often underdosed when treated with TSEBT alone and may require supplemental boosts with appropriately selected electrons or photons to ensure adequate dose delivery at depth.

Limited Superficial Radiotherapy

Minimal IA disease should be treated with a single radiation field where possible, though abutting fields may be required at a convex surface such as the scalp, an axillary fold, breast, hand, or foot. The junction of abutting fields should be shifted during the course of treatment to improve homogeneity. Field margins can be limited by lead cutout to only 1 to 2 cm beyond the visible (or palpable) clinical lesion. Electrons of 6- to 16-MeV energy with appropriate bolus material are sufficient, and the dose should be approximately 30 Gy with the expectation of cure. One or several fields of 6- to 16-MeV electrons can also be used to encompass the limited volumes required to palliate symptomatic skin lesions, including most tumor nodules and skin ulcers. Although a single dose of 4 to 8 Gy may produce a complete response and full symptom relief, a fractionated dose of at least 20 Gy is more likely to achieve a complete and durable clinical remission.^20,^86,¹³⁹ Whether irradiating a limited region of the skin for cure or for palliation, dose prescriptions should specify conventional or lower doses per fraction (e.g., 1.2 to 2 Gy) and the total dose should be limited (e.g., 20 to 30 Gy). This preserves as much radiation tolerance of the skin as possible and enables future delivery of local radiotherapy or TSEBT without moderate-to-severe toxicities. Furthermore, in situations of coexisting cancer diagnoses (e.g., Hodgkin’s lymphoma or breast, prostate, or rectal cancers), radiation treatments for those diagnoses must be carefully planned to minimize the superficial skin exposure in order to preserve the option of TSEBT in the future.

Node, Viscera, and Blood Radiotherapy

Nodal regions (e.g., axillary and inguinal) and some visceral organs (e.g., lungs, larynx, and brain) can be encompassed using standard beam arrangements and energies. Prescriptions are typically 20 to 30 Gy in 10 to 20 fractions over 2 to 4 weeks.

Several experiences using TSEBT in combination with either low-dose, total body irradiation ¹⁷¹ or total nodal irradiation^172,¹⁷³ have been reported. However, such treatment may increase the risk of secondary malignant diseases or myelodysplastic syndrome¹⁷³ and has not demonstrated superiority over conventional systemic agents. For patients undergoing allogeneic bone marrow transplantation, incorporation of total body irradiation in the preparative regimen may be beneficial.¹³⁶ A dose of 12 to 15 Gy in 6 to 10 fractions is recommended; single fractions are discouraged because of a lower therapeutic ratio.

TSEBT delivers a total dose to the blood and marrow of approximately 0.4 Gy in 30 or more fractions over 6 to 14 weeks. In general, hematologic values remain normal during a course of TSEBT. For those with Sézary syndrome, however, the number of circulating Sézary cells may fall by 50% to 95% and, in rare instances, may remain in a sustained remission.

Modern TSEBT

Technical advances since the initial Stanford and St. John’s publications have included dose escalation to 35 to 36 Gy, improvements in shielding, and integration of boost treatments to underdosed regions, culminating in the publication of a consensus statement from the EORTC regarding appropriate techniques for TSEBT ¹⁴³ (Table 77-9).

TABLE 77-9 EORTC Guidelines for Total Skin Electron Beam Therapy

• Dose inhomogeneity in air at treatment distance should be <10% within vertical and lateral dimensions.

• 80% isodose line should be ≥4 mm deep to the skin surface to ensure that the epidermis and dermis fall within the high-dose region.

• 80% isodose line should receive a minimum total dose of 26 Gy.

• 20% isodose line should be <20 mm from the skin surface to minimize dose to underlying structures.

• 30 to 36 fractions should be used to minimize acute side effects.

• Total dose to bone marrow from photon contamination should be <0.7 Gy.

• Patch treatments should be used to underdosed areas, such as the perineum, scalp, and soles of feet.

• Internal and external eye shields should be used to ensure that the dose to the globe is not more than 15% of the prescribed skin surface dose.

Adapted from Jones GW, Kacinski BM, Wilson LD, et al: Total skin electron radiation in the management of mycosis fungoides. Consensus of the European Organization for Research and Treatment of Cancer (EORTC) Cutaneous Lymphoma Project Group. J Am Acad Dermatol 47:364-370, 2002.

The method of TSEBT used clinically at Yale achieves these objectives with a 6-MeV modern linear accelerator using the dual-field, six-treatment position technique. The patient stands 3.8 m from the accelerator head, which rotates 17.5 degrees above and below horizontal to produce the dual fields. A 3.2-mm polycarbonate screen (Lexan) is placed in front of the patient and serves to attenuate and scatter the incident electrons, resulting in an electron energy of 3.9 MeV at the skin surface (Fig. 77-4). A total of six treatment positions are designated: anteroposterior, right and left anterior oblique, posteroanterior, and right and left posterior oblique (Fig. 77-5). These positions maximize skin unfolding, thereby improving dose homogeneity in the lateral dimension. Each position is treated with two fields, an upper field and a lower field, in order to maximize dose homogeneity in the vertical dimension. On treatment day 1, the anteroposterior, right posterior oblique, and left posterior oblique positions are treated. On treatment day 2, the posteroanterior, right anterior oblique, and left anterior oblique positions are treated with the same dose. Over the course of a 2-day treatment cycle, a patient will receive 2 Gy to the entire skin surface. This pattern continues, with patients receiving treatment 4 days a week for a total of 9 weeks, during which time a total dose of 36 Gy is delivered to the skin surface (Table 77-10).

Figure 77-4 Geometry of dual-field total skin electron beam technique (TSEBT).

Figure 77-5 Treatment positions used in total skin electron beam therapy. Top row (from left to right), right anterior oblique, anteroposterior, and left anterior oblique treatment positions. Bottom row (from left to right), right posterior oblique, posteroanterior, and left posterior oblique treatment positions.

Reprinted with permission from Figure 1 in Smith BD, Wilson LD: Management of mycosis fungoides. Part 2. Treatment. Oncology 17:1419-1433, 2003.

TABLE 77-10 Treatment Protocol at Yale–New Haven Hospital

Treatment Cycle
Day 1	AP, RPO, LPO treatment positions
Day 2	PA, RAO, LAO treatment positions
Dose
Dose per cycle	2 Gy
Cycles per week	2
Total cycles	18
Total dose	36 Gy
Boosts
Perineum	100 cGy/day, first 9 and last 9 treatment days
Soles of feet	100 cGy/day, first 7 and last 7 treatment days
Blocking
External eye shields	First 11 cycles
Internal eye shields	Last 7 cycles
Lip shield	Cycles 1-4
Lead mitt for hands	Every other cycle
Fingernail shield	Every other cycle, alternating with mitts
Foot block	Cycles 1-3, 5, 7, 9, 11, 13, 15, 17, 18
Testicular shield	Used with perineal boost only

AP, anteroposterior; LAO, left anterior oblique; LPO, left posterior oblique; PA, posteroanterior; RAO, right anterior oblique; RPO, right posterior oblique.

Data from Chen Z, Agostinelli AG, Wilson LD, et al: Matching the dosimetry characteristics of a dual-field Stanford technique to a customized single-field Stanford technique for total skin electron therapy. Int J Radiat Oncol Biol Phys 59:872-885, 2004.

This technique places the dose maximum at 1 mm, the 80% isodose line at 6 mm, and the 20% isodose line at 12 mm,¹⁷⁴ thereby satisfying the EORTC criteria for TSEBT.¹⁴⁶ Photon contamination due to bremsstrahlung scattering in the machine head, intervening air, scatterers or degraders, and patient, is acceptable at 1.2%. A comparison of the depth-dose curves for a single anteroposterior treatment position and for all six treatment positions is presented graphically in Figure 77-6 and by film dosimetry¹⁷⁴ in Figure 77-7. Using six treatment positions shifts the isodose curve toward the skin surface due to the obliquity of the incident electrons.

Figure 77-6 Depth dose curves for dual-field total skin electron beam therapy (TSEBT). Blue line, Depth dose curve for an anteroposterior dual-field 4-MeV electron beam incident on polystyrene, a tissue-equivalent material. Note that the surface receives approximately 80% and that the maximum dose is deposited 8 mm from the surface. Red line, Depth dose curve for dual-field, six-treatment position, standard TSEBT. Due to multiple beams entering the skin surface at oblique angles, the dose delivered to the skin surface rises dramatically and drops to 80% by 8 mm.

Graph courtesy Zhe Chen, PhD; adapted with permission from Figures 6 and 11 in Chen Z, Agostinelli AG, Wilson LD, et al: Matching the dosimetry characteristics of a dual-field Stanford technique to a customized single-field Stanford technique for total skin electron therapy. Int J Radiat Oncol Biol Phys 59:872-885, 2004.

Figure 77-7 Cross-sectional dose distribution. Axial film dosimetry for single-treatment position and six-treatment position techniques. The dark region is proportional to the electron dose delivered to the film.

Adapted with permission from Figure 5 in Chen Z, Agostinelli AG, Wilson LD, et al: Matching the dosimetry characteristics of a dual-field Stanford technique to a customized single-field Stanford technique for total skin electron therapy. Int J Radiat Oncol Biol Phys 59:872-885, 2004.

Because of shielding inherent to these treatment positions, the soles of the feet, perineum, and scalp are underdosed and always require supplemental treatment. One can use a dose of 120 kV, HVL 4.2 mm Al, to deliver patch treatments to the soles of the feet (14 Gy in 14 fractions on treatment days 1 to 7 and 30 to 36) and perineum (18 Gy in 18 fractions on treatment days 1 to 9 and 28 to 36). The scalp dose can be supplemented with either orthovoltage patches (6 to 20 Gy over 1 to 3 weeks) or with an electron reflector mounted above the patient’s head ¹⁷⁵ (see Table 77-10). Other areas of potential underdosing include the ventral penis, upper medial thighs, inframammary folds, folds under any pannus, and lateral and flatter regions of the face and trunk. Supplemental patch fields, as guided by in vivo dosimetry or clinical suspicion, are appropriate for these regions to ensure that the surface dose is at least 50% of the prescribed TSEBT dose.^176,¹⁷⁷ When determining the total dose given in patch treatments, doses to areas such as the feet and perineum may be reduced as clinically indicated to enhance patient tolerance, provided such areas are uninvolved.

The hands, wrists, ears, ankles, and dorsal penis may be overdosed because of a variety of factors, including tissue heterogeneity (e.g., bone), high convexity, and overlap between more than three of the six primary fields.⁸⁶ These regions should therefore be shielded for a certain proportion of the treatments to reduce the total dose to 36 Gy or less. Detailed dosimetric measurements are typically required to determine the most appropriate shielding regimen for a particular treatment arrangement and patient geometry (Table 77-11). Lead, bags containing rice, and plywood boards are appropriate methods of shielding. The eyes can be shielded with a combination of external eye shields and lead internal eye shields.¹⁷⁸ Although internal eye shields may cause conjunctival irritation and corneal abrasions, the risk is usually less than 1%.¹⁴⁶ The shielding regimen used at the Yale–New Haven Hospital and the resulting doses to various anatomic structures are presented in Tables 77-10 and 77-11.

TABLE 77-11 Dose to Anatomic Sites Using In Vivo TLD Dosimetry

Because many tumors have a thickness greater than 6 mm, TSEBT alone may result in significant underdosing. As a result, for patients with bleeding, weeping, or painful tumors at presentation, an initial or concurrent boost of 10 Gy in five fractions with either 6- to 16-MeV electrons or orthovoltage photons has been effective. In addition, asymptomatic plaques and tumors that persist at the end of treatment receive a similar boost to ensure adequate dose delivery at depth.

Other Radiation Management Issues

Cutaneous symptoms, including xerosis, pruritus, and pain from fissures or ulceration, are often severe and should be managed aggressively with topical steroids, emollients, oral antihistamines and aggressive wound care with nonocclusive dressings.⁷² Acute skin toxicity from TSEBT usually responds to the above measures, although treatment breaks on the order of 1 to 2 weeks may occasionally be required.¹⁷⁹ Large bullae that develop during TSEBT should be lanced under sterile conditions and require nonocclusive dressings until they reepithelialize.

As previously discussed, patients with advanced MF may experience a profound degree of immunosuppression. As a result, cutaneous and systemic infections are an important cause of morbidity and mortality. The most common infections include cellulitis due to Staphylococcus aureus or β-hemolytic streptococci, cutaneous herpes simplex or herpes zoster (both of which may disseminate cutaneously), bacteremia (most commonly with S. aureus), and bacterial pneumonia.³² Signs and symptoms of such infections should be evaluated promptly and treated aggressively. Independent of MF, TSEBT also suppresses cutaneous immunity, likely due to destruction of normal skin-homing lymphocytes and epidermal antigen-presenting dendritic cells. Any new eruption that develops during the course of TSEBT should therefore be evaluated promptly because disseminated cutaneous bacterial, fungal, and viral infections are not uncommon¹⁸⁰ (Fig. 77-8).

Figure 77-8 Disseminated herpes simplex during the course of total skin electron beam therapy (TSEBT). Patient developed a disseminated pustular eruption shortly after beginning TSEBT. Numerous pustules were noted throughout the body, but all were localized to preexisting MF patches and plaques. Direct fluorescent antibody testing was positive for herpes simplex.

From Smith BD, Son CB, Wilson LD: Disseminated herpes simplex after total skin electron beam radiotherapy for mycosis fungoides. J R Soc Med 96:500-501, 2003.

Treatment Algorithm, Controversies, Problems, Challenges, and Future Possibilities

A treatment algorithm for newly diagnosed and recurrent MF ¹⁸¹ is presented in Table 77-12. Multiple treatment options are listed, reflecting different institutional preferences and the paucity of randomized trials comparing different treatment modalities. A major challenge is the need for prospective randomized trials to compare radiation-based and non-radiation-based approaches to initial treatment. Adequately powered trials are problematic given the rarity of MF and the difficulty in accruing patients. In the absence of randomized data, therefore, treatment will continue to be guided by institutional expertise and patient preference.

TABLE 77-12 Treatment Recommendations by Stage

Stage	Initial Treatment	Treatment for Relapsed or Refractory Disease
Unilesional patch T1N0M0 (stage IA)	Localized, superficial radiotherapy Topical BCNU Topical HN2 Topical bexarotene PUVA UVB	Same as initial treatment
Limited patch-plaque T1NO-1MO (stage LA)	Topical HN2 Topical BCNU PUVA UVB (if only patches) Topical bexarotene TSEBT ± PUVA or HN2	Same as initial treatment TSEBT Topical or oral bexarotene Interferon-α alone PUVA + interferon-α Topical corticosteroids
Extensive patch-plaque	Topical HN2	Same as initial treatment
Symptoms controlled	Topical BCNU	TSEBT
Minimal plaque thickness T2N0-1M0 (stage IA, IIA)	PUVA ± interferon-α UVB (if only patches) TSEBT + adjuvant PUVA or HN2	Topical or oral bexarotene Interferon-α alone Topical corticosteroids
Extensive patch-plaque	TSEBT + adjuvant PUVA or HN2	Topical HN2
Symptomatic	—	Topical BCNU
Indurated plaques T2N0-1M0 (stage IB, IIA)	—	PUVA + interferon-α Topical or oral bexarotene Topical corticosteroids Repeat TSEBT
Cutaneous tumors T3N0-1M0 (stage IIB)	TSEBT + adjuvant HN2 or photopheresis Bexarotene Denileukin diftitox	Topical HN2 and local radiotherapy Oral bexarotene Denileukin diftitox Systemic chemotherapy Repeat TSEBT
Erythroderma T4N0-1M0 (stage III)	TSEBT + adjuvant photopheresis PUVA + interferon-α Photopheresis alone Interferon-α	Methotrexate Oral bexarotene Denileukin diftitox Gemcitabine Fludarabine ± interferon-α Pentostatin ± interferon-α 2-chlorodeoxyadenosine Standard B-cell lymphoma chemotherapy Repeat TSEBT Allogeneic transplant
Nodal or visceral disease Stage LV	Clinical trial Allogeneic transplant Methotrexate TSEBT Topical HN2 Topical BCNU PUVA + interferon-α Topical corticosteroids Localized, superficial radiotherapy	Same as initial treatment Systemic chemotherapy as listed
Transformed to large cell variant	Same as stage IV Consider rituximab if CD20+¹ Trimetrexate Allogeneic transplant	Same as stage IV

BCNU, bischloroethylnitrosurea; HN2, mechlorethamine; PUVA, psoralen plus ultraviolet A; TSEBT, total skin electron beam therapy; UVB, ultraviolet B.

Adapted with permission from Smith BD, Wilson LD: Management of mycosis fungoides. Part 2. Treatment. Oncology (Huntingt) 17:1419-1433, 2003 (Table 2).

The clinical management of patients with MF is complex and typically requires an interdisciplinary team approach that attends to the patient as a whole. Most patients require a fairly detailed understanding of the disease, its implications and risks, and all treatment options. The majority of patients are seeking a cure, which for some is unrealistic. Improvements in quality of life are possible regardless of presenting stage. The relative contributions of various factors in determining quality of life have not been systematically explored. Clinical experience indicates, however, that the most salient factors are controlling symptoms; achieving remission regardless of prior clinical trajectory; using treatments that are efficient and of limited duration (so as to be less disruptive to lifestyle and personal circumstances over the long term); addressing any anxiety and depression; and providing sustained and reliable support and expertise that extends into years and even decades of follow-up.

Many biologic and clinical questions pertaining to the pathophysiology and treatment of MF remain unanswered. For example, it is unclear whether the development of MF is initially driven by a specific exogenous antigen or by random mutational events occurring in activated T cells. In addition, the molecular factors that influence disease progression and response to various treatment modalities are just beginning to be elucidated. Clinically, the staging system is undergoing major revision, and more studies with homogeneous patient populations are required to evaluate putative prognostic factors.

The future will likely bring additional novel systemic therapies for MF. For example, biologic and immunologic agents including monoclonal antibodies directed against CD4 and CD52 (alemtuzumab, Campath-1H), topical tazorotene (a novel retinoid), interleukin-2, interleukin-12, and individualized dendritic cell-based tumor vaccines are under investigation and may continue to improve the treatment of recurrent or refractory disease.^182,¹⁸³ Strategies for integrating novel biologic agents with standard therapies such as TSEBT are needed to maximize the benefits of each. For those with advanced disease, a combination of TSEBT to induce cutaneous remission and biologic therapy to induce systemic remission shows the greatest promise for providing long-term disease control.

Critical References

15 Wood GS, Salvekar A, Schaffer J, et al. Evidence against a role for human T-cell lymphotrophic virus type I (HTLV-I) in the pathogenesis of American cutaneous T-cell lymphoma. J Invest Dermatol. 1996;107:301-307.

16 Li G, Vowels BR, Benoit BM, et al. Failure to detect human T-lymphotropic virus type-I proviral DNA in cell lines and tissues from patients with cutaneous T-cell lymphoma. J Invest Dermatol. 1996;107:308-313.

31 van Doorn R, Dijkman R, Vermeer MH, et al. A novel splice variant of the Fas gene in patients with cutaneous T-cell lymphoma. Cancer Res. 2002;62:5389-5392.

37 Wasik MA, Vonderheid EC, Bigler RD, et al. Increased serum concentration of the soluble interleukin-2 receptor in cutaneous T-cell lymphoma. Clinical and prognostic implications. Arch Dermatol. 1996;32:42-47.

49 Diamandidou E, Colome-Grimmer M, Fayad L, et al. Transformation of mycosis fungoides/Sezary syndrome. Clinical characteristics and prognosis. Blood. 1998;92:1150-1159.

50 Vergier B, de Muret A, Beylot-Barry M, et al. Transformation of mycosis fungoides. Clinicopathological and prognostic features of 45 cases. French Study Group of Cutaneous Lymphomas. Blood. 2000;95:2212-2218.

53 Stein M, Farrar N, Jones GW, et al. Central neurologic involvement in mycosis fungoides. Ten cases, actuarial risk assessment, and predictive factors. Cancer J. 2006;12:55-62.

58 LeBoit PE. Granulomatous slack skin. Dermatol Clin. 1994;12:375-389.