Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Mycosis fungoides and Sézary syndrome are two closely related conditions in which neoplastic T cells infiltrate the skin and circulate in the peripheral blood. Both conditions are neoplasms that typically have a mature helper T-cell phenotype and a propensity to colonize the epidermis. Because individual patients can have discrete cutaneous lesions at one point in time and erythroderma with circulating neoplastic cells at another time, some advocate the term cutaneous T-cell lymphoma to describe what they consider to be a single disease. However, the delineation of a variety of other distinct clinicopathologic entities that are also cutaneous T-cell lymphomas has, in our opinion, rendered this term imprecise and obsolete. For example, the skin may be the only site involved by anaplastic large T-cell lymphoma.
This chapter covers mycosis fungoides and its many variants and Sézary syndrome. Other primary cutaneous T-cell lymphoproliferative disorders are considered elsewhere in this text.
Mycosis fungoides is a T-cell lymphoma in which lymphocytes infiltrate the epidermis in its early stages, resulting in flat, often slightly scaly lesions (patches). In some patients, lymphocytes acquire the ability to proliferate in the dermis, forming plaques and nodules. A small minority of patients will have involvement of internal organs in the course of their disease. Most cases of mycosis fungoides have a T-helper phenotype, but clinically and histopathologically identical infiltrates can be seen in which T suppressor cells or even B cells are present. Our view, which is not shared by all authorities, is that the clinical evolution of patches to plaques and tumors is what determines whether a patient has the disease mycosis fungoides, not a specific immunophenotype. If a patient has indolent patches in which there are epidermotropic CD8 + T cells, there is no harm in labeling that mycosis fungoides. Indeed, in many centers, immunophenotypic studies are not routinely performed, and patients are treated with excellent results. The term mycosis fungoides does not apply, however, to a disease caused by infection with the retrovirus human T-lymphotropic virus 1 (HTLV-1), despite the clinical and pathologic resemblance of some cases; that condition is referred to as adult T-cell leukemia/lymphoma . The major diagnostic features of mycosis fungoides are listed in Box 39-1 .
Patch stage: Patches with fine overlying scale, often more than 5 cm in diameter in photoprotected sites
Plaque stage: Same location and size as patches, but lesions are thicker with induration and elevation
Tumor stage: Solid nodules at least 1 cm in diameter that usually develop in patches and plaques
Patch stage: Sparse perivascular to bandlike infiltrate of lymphocytes with variable infiltration of the epidermis and variable cytologic atypia
Plaque stage: Infiltrate denser and extends into reticular dermis
Tumor stage: Diffuse infiltrate extending throughout the reticular dermis
βF1 + , CD3 + , CD4 + , CD8 − immunophenotype is most common, but variations in otherwise typical disease occur and have little meaning
Clonality common but not obligatory by PCR-based gamma alone or gamma plus beta chain gene rearrangements
Mycosis fungoides is largely a disease of middle-aged and older people. However, as clinicians and pathologists have become more adept at recognizing its early stages, more cases in young adults and even in children have come to light. The incidence of mycosis fungoides in a population is certainly affected by the number of dermatologists in the community, their interest in and awareness of the disease, and their threshold for diagnosis. An interesting observation is that the incidence of mycosis fungoides rose rapidly in the early 1980s, coincident with the delineation of criteria for the diagnosis of patch-stage disease by Sanchez and Ackerman. After the publication of their paper, many pathologists began to diagnose mycosis fungoides on the basis of infiltrates they might have previously regarded as parapsoriasis en plaques or inflammatory conditions, such as spongiotic dermatitis. The increased incidence of mycosis fungoides in the United States seems to reflect a rise in the detection and diagnosis of early patch-stage disease.
A number of investigators have tried to link mycosis fungoides to environmental exposures, without success. Studies to determine whether common inflammatory skin diseases, such as atopic dermatitis, chronic allergic contact dermatitis, and psoriasis, give rise to mycosis fungoides are undermined by several factors. Early patches of mycosis fungoides can resemble these diseases clinically, so a patient with a 20-year history of “atopic dermatitis” preceding mycosis fungoides might have had patches of mycosis fungoides that were simply not recognized as such. Early patch-stage lesions of mycosis fungoides can resemble various inflammatory conditions under the microscope, so that even “biopsy-proven” psoriasis may not be that disease at all.
Several studies have sought the presence of a virus in the cells of mycosis fungoides. In the 1970s, interest centered on the identification of viral particles in skin biopsy samples of mycosis fungoides by electron microscopy. More recently, interest has focused on a possible role for HTLV-1, the retrovirus that causes adult T-cell leukemia/lymphoma, in mycosis fungoides. An initial study seemed to identify partial viral transcripts in the cells of mycosis fungoides, but further investigation has not borne this out in most cases. Another theory related to infection is that mycosis fungoides is an abnormal response to bacterial superantigens.
Mycosis fungoides is largely defined by the clinical features of its early stages. Requisite to this definition is an initial presentation as flat, scaly lesions called patches. These first appear in areas of the skin that are best protected from sunlight—the buttocks and groins of both sexes and the breasts of women. Subtle wrinkling, slight erythema, telangiectasias, and either hypopigmentation or hyperpigmentation are variable findings. Often, the patches are so subtle that patients do not notice them for some time, and both patients and their physicians may attribute the condition to dry skin or atopic dermatitis. Patches are generally round or oval, although they are sometimes finger shaped or digitate ( Fig. 39-1 ). Their size can range from about 1 cm to more than 15 cm.
Some clinicians use the term small plaque parapsoriasis to refer to small patches (smaller than an adult palm) and the term large plaque parapsoriasis to refer to larger lesions of patch-stage mycosis fungoides. Those who think that mycosis fungoides begins as an inflammatory condition that may regress often use the term parapsoriasis . This usage is based on the work of the French dermatologist Brocq in the late 19th and early 20th centuries. He envisaged a complex relationship among psoriasis, eczema, seborrheic dermatitis, the conditions now known as pityriasis lichenoides acuta and chronica, and mycosis fungoides. In our opinion, the term parapsoriasis is invalid scientifically, although it may have some functional utility in that it is shorthand for “I suspect that this is an early patch of mycosis fungoides but am not sure.” This dilemma is better expressed in clear language, however. Confounding this already confusing situation is the habit of some dermatologists to use parapsoriasis to refer to pityriasis lichenoides, an inflammatory disease.
Digitate or finger-shaped patches may occur by themselves or with the conventional lesions of mycosis fungoides. This has led some observers to conclude that the condition formerly called digitate dermatosis is in fact a form of mycosis fungoides. Because the prognosis of patients with only digitate lesions is excellent and the histopathologic findings are often paltry, others question the usefulness of labeling patients with digitate dermatosis as having mycosis fungoides. The first documented report of a patient with digitate dermatosis evolving into conventional mycosis fungoides did not appear until recently, pointing out the importance of at least recognizing that patients with this condition have a different prognosis.
In some patients with pre-existing patches of mycosis fungoides and in others who claim that they never had such patches, areas of the skin can become thin and wrinkled and marked by macules of hypopigmentation and hyperpigmentation, along with telangiectasias. This appearance is known as poikiloderma or poikiloderma vasculare atrophicans . It appears to be a manifestation of regression of patch-stage mycosis fungoides.
The large majority of patients with mycosis fungoides who have patches over a small area of skin at presentation prove to have an indolent condition that seldom becomes more than a cosmetic problem, even if it is untreated. In a minority of such patients, disseminated patches arise.
Plaques of mycosis fungoides are usually located in the same locations as patches but differ from patches in that they are raised. Plaques are varying shades of red to red-brown and scaly. They are often polycyclic, with clearing in the center ( Fig. 39-2, A ). They sometimes ulcerate, but not as much as nodules or tumors do.
Tumors of mycosis fungoides are raised nodules that are often smooth but frequently ulcerate ( Fig. 39-2, B ). They are clinically indistinguishable from nodules and tumors of other cutaneous lymphomas. However, tumors almost invariably arise within or adjacent to pre-existing patches and plaques of mycosis fungoides, and thus a careful clinical examination to assess for concomitant patches and plaques can be helpful to distinguish tumors of mycosis fungoides from other cutaneous lymphomas. The tumors can sometimes assume a mushroom-like configuration, and this attribute resulted in Alibert's giving the condition the name mycosis fungoides .
There is vast variability in the histopathologic appearance of mycosis fungoides, especially in patch-stage disease. This reflects the fact that early lesions may be composed largely of non-neoplastic cells, exerting their influence through cytotoxicity and cytokine production and in ways not yet appreciated.
The early patches of mycosis fungoides feature lymphocytes that are not usually morphologically abnormal and thus can appear similar to the lymphocytes found in inflammatory skin diseases. In fact, studies have demonstrated that only 4% (27 of 745) of biopsy specimens of early mycosis fungoides reviewed had atypical lymphocytes within the epidermis. Therefore, identification of a section as representing early mycosis fungoides, either definitely or possibly, usually requires attention to the histopathologic pattern of the infiltrate rather than the identification of atypical lymphocytes.
The earliest patches of mycosis fungoides feature small lymphocytes around venules of the superficial plexus; some are scattered interstitially in the papillary dermis, with only a few within the epidermis ( Fig. 39-3 ). In some cases, when the cells of mycosis fungoides enter the epidermis, they can elicit spongiosis or edema between keratinocytes. The degree of spongiosis is usually less than that seen when the same number of lymphocytes enter the epidermis in an inflammatory skin disease. The tendency of the cells of mycosis fungoides to colonize the epidermis is referred to as epidermotropism . This term is also used to connote that there are areas of the epidermis that have only slight spongiosis and many lymphocytes. Exocytosis describes the migration of inflammatory cells into the epidermis and is a more neutral term. Because the term epidermotropism presupposes the ultimate diagnosis, it is best avoided if the diagnosis of mycosis fungoides is equivocal.
In early patch-stage disease, mycosis fungoides is often not recognizable with certainty. As the patches develop, the papillary dermis becomes fibrotic. The collagen bundles of the papillary dermis are usually fine and haphazardly oriented. This meshwork changes to one in which there are coarse fibers sometimes likened to “pink fettuccini.” At the same time, rete ridges begin to elongate, usually only slightly and very evenly. Their bases remain rounded, unlike in many interface dermatitides. Lymphocytes may lodge in the basal layer of the epidermis, with only slight vacuolar changes and few necrotic keratinocytes.
The papillary dermal lymphocytic infiltrate often becomes bandlike, at least in foci. The combination of elongated rete ridges with rounded bases and bandlike lymphocytic infiltrates is known as a psoriasiform lichenoid pattern ; if spongiosis is also present, it is referred to as a spongiotic psoriasiform lichenoid pattern . If the lymphocytes engaged as a host response to the neoplasm kill keratinocytes that constitute rete ridges, the epidermis may become thin and flat based—an atrophic lichenoid pattern . These three patterns should raise the pathologist's suspicion that he or she may be dealing with a lesion of mycosis fungoides because only a few inflammatory skin diseases share these patterns ( Box 39-2 ).
Mycosis fungoides, patch stage
Secondary syphilis (usually superficial and deep, with many plasma cells and histiocytes)
Lichenoid purpura (extravasated erythrocytes and siderophages)
Lichen striatus (linear eruption of papules in a child or teenager)
Early lesions of lichen sclerosus et atrophicus
Surface of some lesions of morphea
Drug reaction (one pattern among many)
Mycosis fungoides, patch stage
Urticarial stage of bullous pemphigoid
Drug reactions (one pattern among many)
Allergic contact dermatitis (rare; so-called lichenoid contact dermatitis)
Chronic photoallergic dermatitis (actinic reticuloid)
Mycosis fungoides, atrophic patch stage
Atrophic lichen planus
Lichenoid purpura
Regression of melanoma, Bowen's disease, superficial basal cell carcinoma
Centers of lesions of porokeratosis (sometimes)
Poikilodermatous lesions of dermatomyositis
As the infiltrates of mycosis fungoides become dense and bandlike in the papillary dermis, they also begin to exhibit cells with atypical nuclei ( Fig. 39-4 ). Cells of patch-stage mycosis fungoides have slightly larger nuclei than those of lymphocytes in inflammatory conditions, with an irregular nuclear contour—the so-called cerebriform lymphocyte ( Fig. 39-5 ). An important caveat is that if nuclear atypia is used as a criterion for the differential diagnosis between a patch of mycosis fungoides and an inflammatory condition, the atypia must be unmistakable. Many pathologists can convince themselves that the nuclei of lymphocytes are atypical by staring at them for too long under an oil immersion lens.
Some patches of mycosis fungoides feature epidermal atrophy, in concert with a patchy lichenoid lymphocytic infiltrate. The papillary dermis is often markedly fibrotic and contains telangiectasias and melanophages, corresponding to the clinical picture of poikiloderma vasculare atrophicans. In such atrophic patch-stage lesions, it may be difficult to demonstrate a sufficient number of lymphocytes in the epidermis to rule out an inflammatory disease with an atrophic lichenoid pattern ( Box 39-2 ).
In plaques of mycosis fungoides, lymphocytes extend into the reticular dermis, not only around vessels but also interspersed between reticular dermal collagen bundles ( Fig. 39-6 ). This finding occurs beneath an epidermis and papillary dermis displaying the changes described earlier for fully developed patches of mycosis fungoides. Although lymphocytes with atypical nuclei are few in early patches and more numerous in late ones, they almost always constitute a significant percentage of the infiltrate in plaques. Similarly, aggregations of lymphocytes, termed Pautrier's microabscesses or collections, are rare in patches but common in plaques. (Interestingly, this distinctive clue to the diagnosis of mycosis fungoides was discovered not by Pautrier but by Darier. ) The atypical lymphocytes of patches have scant cytoplasm and irregular, sometimes cerebriform lymphocytes. By contrast, in plaques, many of the lesional lymphocytes often have large vesicular nuclei, large nucleoli, and some discernible cytoplasm. Furthermore, in contrast to patches, which lack eosinophils and plasma cells, plaques and tumors of mycosis fungoides often have many of these cells. This might correlate with a shift from Th1 to Th2-like cytokine production as lesions change from patches to plaques.
Nodules or tumors of mycosis fungoides acquire their clinical features by virtue of lymphocytic infiltrates that are present as nodules or diffusely replace the reticular dermis ( Fig. 39-7 ). Large cell transformation, defined as when more than 25% of the infiltrate is composed of large cells, can occur in tumors of mycosis fungoides. The appearance of the large cells ranges from cells with large round or slightly oval vesicular nuclei and scant cytoplasm to cells with large oval vesicular nuclei, large nucleoli, and abundant cytoplasm, similar to the cells of anaplastic large cell lymphoma. This usually occurs in advanced disease and may have an adverse prognostic impact. Nevertheless, large cell transformation can occasionally occur in patches and plaques as well, although less commonly than in tumors ( Fig. 39-8 ).
Anaplastic large cells may predominate to such an extent that only the clinical identification of patches or plaques at other sites allows the distinction from anaplastic large cell lymphoma. Although lymphocytes home to the epidermis in patch- and plaque-stage lesions, some tumors of mycosis fungoides completely lack intraepidermal lymphocytes. The loss of dependence on an epidermal environment for cellular proliferation in the skin occurs apace with the cells' capacity to lodge in internal organs in mycosis fungoides.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here