Primary Cutaneous T-Cell Lymphoma: Review and Current Concepts
Primary cutaneous T-cell lymphomas (CTCLs) encompass a wide variety of lymphomas that are characterized by the localization of the malignant lymphocytes to the skin at presentation. Advances in molecular biologic techniques, including immunophenotyping and gene rearrangement studies to determine clonality, have led to more frequent diagnosis of CTCL as well as more consistent subclassification of these entities. However, there continues to be confusion in the classification, prognosis, and management of patients with CTCL. The purpose of this review is to present a summary of the diagnosis, prognosis, and treatment of CTCL, with specific emphasis on mycosis fungoides (MF) and Sézary syndrome (SS). We also present a detailed discussion of the entities that make up the differential diagnosis of CTCL.
MF and SS are the most common forms of CTCL. The etiology of this disease is still unknown. Patients may go for months to years with skin abnormalities before being diagnosed. MF/SS is an indolent disease and patients with T1 disease have a normal life expectancy. Patients who undergo transformation to large-cell lymphoma (8% to 23% of patients) have a poor prognosis, with mean survival ranging from 2 to 19 months. Treatment for MF/SS continues to be palliative. There are many new therapies that are currently being investigated in clinical trials, and the DAB389IL-2 fusion protein was recently approved for the treatment of refractory MF/SS.
CUTANEOUS T-CELL lymphomas (CTCL) are a group of lymphoproliferative disorders characterized by localization of the neoplastic T lymphocytes to the skin at presentation. Although mycosis fungoides (MF) and Sézary syndrome (SS) make up the great majority of these diseases, there are many distinct entities that make up the differential diagnosis of cutaneous lymphoproliferative disorders. These entities can have quite dissimilar presentations, pathologic findings, prognoses, and treatments. It is imperative that clinicians appreciate these differences to provide appropriate management of these patients. Until recently, classification systems for lymphomas did not make the distinction between primary and secondary cutaneous lymphomas, and most classifications have been based only on histologic diagnosis. The recently published European Organization for Research and Treatment of Cancer classification for primary cutaneous lymphomas may be the most clinically relevant classification. The correct diagnosis of primary cutaneous lymphoproliferative disorders requires evaluation of the clinical presentation, histology, immunophenotyping, and gene rearrangement studies.
EORTC Classification for Primary Cutaneous Lymphomas:
In this review, we present a detailed description of MF/SS, by far the most common form of CTCL. This is followed by descriptions of the other CTCLs, as well as presentations of the many diseases that make up the differential diagnosis of CTCL.
MF and SS make up the great majority of cases of primary CTCL; therefore, a significant portion of this article will be a detailed review of these entities to illustrate concepts to the reader.
Plaque lesions are more erythematous than patch lesions, have well-demarcated margins, and scaling. They can arise in areas of previous patches or arise de novo. In MF, compared with other CTCLs, tumors are frequently found associated with previous patches or plaques, and in these cases, represent loss of epidermotropism with extension of the infiltrate into the deep dermis. Ulceration, with secondary infection of the tumors, is a common cause of morbidity. The D’emblee presentation signifies the appearance of the tumor stage of disease without antecedent patch or plaque disease.
SS accounts for 5% of new presentations of MF/SS. The clinical picture includes generalized erythroderma, pruritus, and circulating malignant cells (Sézary cells). There is some argument over the minimum number of atypical cells needed in the peripheral blood to make this diagnosis. This is usually expressed as an absolute number of atypical cells ( 1,000/mL) or a percentage of the total lymphocyte count (5% to 20%).
Nuclear contour index is a value that can be determined by electron microscopic analysis of the degree of lymphocyte nuclear folding in skin biopsies. The nuclear contour index can differentiate MF from benign lymphoid infiltrates.
Immunophenotyping is performed on the skin biopsies. The malignant cells are usually CD3+, CD4+, CD8-, CD30-, and CD45RO+. CD7 is often deleted from early lesions, and this may predict a worse prognosis. Advanced MF/SS and some variants may lose expression of CD2, CD5, and sometimes CD4.Routine evaluation now usually includes DNA analysis of the T-cell–receptor gene rearrangement to define the clonal population. This can be performed by Southern blotting, which is 90% to 95% sensitive and analyzes the ß-chain gene rearrangement. Most T cells express the ß receptor, whereas few cells express the receptor. PCR, which is nearly 100% sensitive, is used to analyze the chain, which is rearranged in all T cells, whether they express the ß or receptor.
Lymph node biopsies can be performed at the initial staging in advanced disease, or if nodes are found to be enlarged on physical examination, or by imaging studies. Histopathology of lymph node involvement initially affects the paracortical regions. As progression occurs, there are small to large clusters of atypical cells with preserved nodal architecture, followed by partial or total effacement of the node. Immunophenotyping and T-cell–receptor gene rearrangement analysis is helpful in evaluation of lymph nodes.
Peripheral-blood involvement is demonstrable at all stages of disease because of the normal lymphatic drainage via the thoracic duct. It is much more prevalent in tumor stage disease or SS. The level of circulating malignant cells is inversely correlated with prognosis and should be evaluated at initial presentation and routinely during follow-up. Visceral involvement occurs late in the progression of MF/SS.
Formal staging is according to the tumor-node-metastasis classification system. In a study of 43 patients, Martí evaluated 18 clinical and pathologic predictors of prognosis. They concluded that in a clinical evaluation alone, the T stage and serum lactate dehydrogenase level were the major prognostic factors. In a clinicopathologic evaluation, T stage and thickness of the cutaneous infiltrate measured from the granular layer were major prognostic factors.
The goals of treatment with established therapies are symptom relief and cosmetic improvement because cures are generally unattainable with standard therapies. Kaye evaluated the use of early, aggressive therapy with total-skin electron beam therapy (TSEBT) and chemotherapy versus sequential topical therapies. Although the aggressively treated patients had a higher complete remission rate, this group also had significantly more toxicity. More importantly, there were no significant durable differences between disease-free and overall survival between the two groups.
Treatment is typically given three times a week during the clearing phase, and this is followed by a tapering maintenance phase ranging from once a week to once every 3 to 4 weeks. The dose of 8-MOP is 0.6 mg/kg ingested 2 hours before the UVA exposure. Early studies demonstrated high remission rates for early-stage disease. The Scandinavian study group reported a 58% complete response rate within 2 to 3 months of initiation of treatment. The duration of remission was up to 53 months with maintenance therapy. They also reported an 83% remission rate in tumor stage disease. Herrmann reported Northwestern University’s experience of PUVA therapy, including long-term follow-up. In this study, 82 patients with mostly stage I disease were treated with PUVA. Ninety-five percent of patients had at least a partial response, and 65% had a complete response. The mean duration of complete response was 43 months. The mean survival in this group was 8.5 years.PUVA therapy is usually restricted to patients with stage I and II disease.
Side effects to PUVA are well-tolerated and include nausea caused by psoralen ingestion and actinic effects of UVA exposure, including erythema and pruritus in the short term and chronic dry skin and secondary skin malignancies in the long term. There has recently been a case report published linking PUVA therapy to myelodysplasia and subsequently to acute myelogenous leukemia.
The major side effects of topical NM include irritant or allergic contact dermatitis and secondary squamous cell and basal cell carcinomas. Delayed hypersensitivity reactions occur in up to 58% of patients, requiring therapy to be held at least temporarily and desensitization. Ointment-based NM has been associated with less cutaneous sensitivity; Price reported that among 12 patients who had previous hypersensitivity to aqueous-based NM, only three developed contact dermatitis when subsequently treated with ointment-based NM.
In a long-term follow-up study by Vonderheid 35 (11%) out of 331 patients treated with topical NM had complete remissions lasting more than 8 years. The incidence of squamous cell carcinoma of the skin was 8.6 times that of the general population, and the incidence of basal cell carcinoma of the skin was 1.8 times that of the general population. Analysis of this study is complicated by the fact that many patients were also treated with other modalities, including local radiation, TSEBT, ultraviolet phototherapy, and systemic chemotherapy.
Topical BCNU has demonstrated similar effectiveness as NM. BCNU 10 mg is dissolved in ethanol and then mixed with 60 mL of water. The resulting solution is applied once a day to the total body-surface area, except uninvolved areas of the face, hands, genitals, and skin folds. Zacheim has reported a 92% overall response rate at 3 years among 89 patients with T1 disease and a 64% overall response rate at 3 years among 83 patients with T2 disease. The 5-year survival in these two groups was 97% and 79%, respectively. The median time to achieve a complete response was 9 weeks for patients with T1 disease and 12 weeks for those with T2 disease. Hypersensitivity reactions to BCNU are rare. The most common adverse effect of topical BCNU is erythema, which can be followed by the appearance of telangiectasias. Secondary skin cancers are not usually observed. In the above study, myelosuppression was rare at the standard 10-mg dose and was seen in less than 10% of patients treated at double the standard dose.
Electron beam therapy is better tolerated than external-beam, and is very effective. Electrons are delivered at an energy of 4 to 9 MeV to treat the skin only. Less than 5% of the dose travels beyond 2 cm, and the target volume for most patients does not exceed a depth of 5 mm; therefore, deep organ toxicity such as myelosuppression does not occur. Reddy showed that patients who receive a total dose of greater than 20 Gy have a better 5-year survival and longer duration of complete response than those who receive 20 Gy. The standard total dose for a course of TSEBT is 36 Gy. The electrons are generated by a linear accelerator and commonly administered using dual fixed-angle six-field methods.
TSEBT results in a 56% to 96% complete response rate in patients with stage IA to IIA disease.If no adjuvant therapy is given, there is a relatively high relapse rate. The relapse-free survival rate for stage IA disease ranges from 33% to 52% at 10 years; whereas the relapse-free survival rate is only 16% for patients with stage IB disease or worse.Acute toxicity of TSEBT consists of an epithelial reaction with erythema, pain, and swelling. Hair and nail loss is also seen, as well as loss of sweat gland function, but these are reversible. A 10-day break halfway through the treatment period results in less severe skin reactions. Telangiectasias and xerosis are the most common chronic adverse effects. Skin cancers can occur, and the risk of basal cell and squamous cell carcinoma of the skin is increased in patients who subsequently receive PUVA or topical chemotherapy.There is a theoretical risk of ocular complications because it has been shown that standard eye shields are less effective with electron beam compared with orthovoltage radiation. Also, although myelosuppression is extremely rare with TSEBT, it is possible if the source is contaminated with orthovoltage radiation and local therapy is given to a site of hematopoiesis.
TSEBT is most commonly used in patients with diffuse skin involvement of thick plaques and tumors and in patients with SS. It can also be used for lesions that have been refractory to other skin-directed therapy. Current studies are evaluating whether adjuvant therapy improves survival compared with TSEBT alone. In an evaluation of 14 patients with T1 and T2 disease who received PUVA therapy after TSEBT, Quiros demonstrated significant improvement in disease-free survival in the PUVA group and a trend toward improved overall survival.
Photopheresis is initially performed on 2 consecutive days once a month until maximal clearing has occurred. This is followed by an additional 6 months of monthly therapy and then gradually tapered to 2-month intervals and discontinuation. Edelson reported a 73% response rate in patients with CTCL refractory to previous therapies, most of whom had SS. Patients with SS, near normal CD8+ lymphocyte counts, and a relatively short duration of advanced disease were most likely to respond. Edelson have subsequently reported a median survival of 60 months in this initial cohort of patients. Genotypic evidence of clonality was not required in that study. Fraser-Andrews retrospectively compared a group of patients treated with photopheresis with a group not treated with photopheresis. Inclusion in their study required genotypic evidence of a peripheral-blood T-cell clone. Median survival in the patients treated with photopheresis was only 39 months and was not significantly better than the group not treated with photopheresis. There is now an investigational form of liquid 8-MOP available that is administered directly into the photopheresis machine, so patients do not have to ingest the psoralen.
Combination chemotherapy has also been used for patients with advanced disease. Alkylating agents are used in combination with doxorubicin and/or vinca alkaloids. Although response rates of 80% to 100% are seen, there are no significant survival benefits.
Many recent studies have focused on the use of the purine analogs; 2-deoxycoformycin, 2-chlorodeoxyadenosine, and fludarabine.These agents do not have a single mechanism of action, but all of them interfere with intracellular regulation of deoxyribonucleotide pools. The response rate in studies performed to date have been higher with 2-deoxycoformycin (54% to 100%) than with the other two compounds. The major toxicities of these drugs are neurotoxicity and prolonged immunosuppression, leading to opportunistic infections.
Acute adverse effects to IFN include flu-like symptoms such as fevers, chills, and myalgias. These are usually transient. Chronic adverse effects include malaise, depression, anorexia, and weight loss, which often lead to dose-modification or cessation of therapy. Laboratory abnormalities include leukopenia, thrombocytopenia, and abnormal liver function tests. The leukopenia is mild, dose-related, and reversible. Hypothyroidism can also occur as a side effect of IFN.
Recently, a case report was published describing a case of acrocyanosis associated with elevated antinuclear antibody titers, which was attributed to IFN therapy for MF. Neutralizing IFN antibodies can develop during treatment. In one study, seven (41%) out of 17 patients developed IFN-binding antibodies, and two of these neutralized the function of IFN as demonstrated by a functional assay. The presence of anti-IFN antibodies correlated with diminished clinical response. However, the use of these assays to guide therapy is not reliable. The mechanism of resistance to IFN in an IFN-resistant CTCL cell line has been shown to be lack of STAT1 expression, thus preventing normal signaling via the JAK/STAT pathway.
In a small study comparing chemotherapy with chemotherapy plus etretinate in patients with advanced disease, there was no difference in response rate. The etretinate group did have a median response duration of 6 months compared with 3 months in the chemotherapy-only group.
IL-2. Marolleau published a series of seven patients with CTCL (six with MF/SS) that were treated with high doses of IL-2. Treatment consisted of an induction phase consisting of a 5-day continuous infusion given every 2 weeks for three cycles at a dose of 20,000,000 IU/m2/d and a consolidation phase consisting of 5 monthly cycles of 2-day infusions. The response rate in the MF/SS group was 67%, with two complete responses and two partial responses. Similar to IFN, the most common side effects of IL-2 are fatigue, flu-like symptoms, and depression.
IL-12. As previously mentioned, MF/SS seems to be a disease mediated by TH2 helper cells. Rook have shown that peripheral-blood mononuclear cells (PBMCs) from patients with MF/SS are deficient in the production of IFN and IL-12, which is consistent with the cytokine milieu seen in this disease. Rook have also shown that IL-12 is a potent stimulus for IFN production in in vitro studies using PBMCs from patients with SS. This has important therapeutic implications because the revived capacity of IFN production could potentially reverse many of the immunologic abnormalities in patients with MF/SS. IL-12 may also be important in the therapeutic benefit of retinoids, and it also stimulates cell-mediated cytotoxicity in PBMCs from SS patients.
Other investigational agents. Knox published the results of a phase I trial of chimeric anti-CD4 monoclonal antibody in which seven of eight patients with MF had at least a minor response, with an average freedom from progression of 25 weeks. Horikoshi reported one case of a woman with refractory MF who had a complete response of at least 6 years after administration of IFN. In a phase II trial, five (31%) of 16 patients responded to IFN. In a phase II study of CAMPATH-1H, a monoclonal antibody directed against CD52, four out of eight patients with MF had significant responses, including two complete responses. Also, an anti-idiotype vaccine has been tested in a few patients, with a major response in one patient. Hypericin is a photodynamic compound that is activated by UVA or white light, and in vitro studies have shown that it inhibits the proliferation of neoplastic T cells in patients with CTCL. If effective in trials, treatment with hypericin may prove to be less toxic than standard PUVA because of the ability to use visible light.
Primary Cutaneous Large-Cell Lymphoma
Pleomorphic and Immunoblastic CTCL
Subcutaneous Panniculitis-Like T-Cell Lymphoma
RELATED CONDITIONS AND LOW-GRADE MALIGNANT VARIANTS
Small-Plaque (digitate) Parapsoriasis (SPP)
Histopathology of SPP reveals spongiosis with focal areas of hyperkeratosis, parakeratosis, and exocytosis. The dermis contains a mild superficial perivascular, lymphohistiocytic infiltrate, and dermal edema. There is no progression of histologic features with time.
SPP patients should be reassessed with follow-up examinations every 6 months to every year to confirm the clinical impression of SPP. Patients can be treated with lubrication, topical steroids, UVB irradiation, PUVA, and/or topical NM.
Mucinosis (Alopecia Mucinosa)
The disseminated form of pagetoid reticulosis presents with generalized lesions of the same appearance as in Woringer-Kolopp disease, and the course can be more aggressive. Some patients die of this disease in months, whereas others have normal life expectancy. There is little data available regarding the treatment of disseminated pagetoid reticulosis, although there is evidence that PUVA is effective.
Granulomatous Slack Skin (GSS)
In approximately half of the cases, GSS is associated with other lymphoproliferative malignancies, with the malignancy being diagnosed either before or subsequent to the diagnosis of GSS. MF and Hodgkin’s disease are the most common associated malignancies. Only 34 cases of GSS have been described in the literature, so there is no standard therapy. Many treatment modalities have been tried in patients with GSS, including PUVA, topical and systemic corticosteroids, IFN, multiagent systemic chemotherapy, retinoids (etretinate), and surgical excision. However, none of these treatments have been successful in attaining complete remissions.118
Lymphomatoid Papulosis (LyP)
Clinically, the appearance of the lesions in LyP varies. They usually appear on the trunk and extremities. Initially, scattered erythematous papules form, which often progress to brown, hemorrhagic, and/or necrotic lesions over days to weeks. These lesions usually ulcerate, and eventually resolve spontaneously, often leaving scars. The entire course for a cluster of lesions usually lasts 2 to 8 weeks. The intensity and frequency of the outbreaks is variable.
Histopathology of LyP lesions reveals a superficial and deep dermal infiltrate that is both perivascular and interstitial and contains a heterogeneous population of lymphocytes, histiocytes, eosinophils, and atypical small- and large-sized lymphoid cells. LyP is commonly divided into two variants, although these variants often overlap. Type A lesions consist of large cells, which can resemble Reed-Sternberg cells, whereas type B lesions are made up of smaller lymphoid cells that are similar to the atypical cells described in MF. Atypical lymphocytes stain for CD30 and are usually positive for HLA-DR and CD25. T-cell receptor analysis often demonstrates clonality.
The diagnosis of LyP versus CD30+ primary cutaneous lymphoma is often difficult because of the similar clinical and histologic presentation. The definition of LyP is a natural history that is benign, with chronic outbreaks that spontaneously resolve. Therefore, a period of observation may be required in difficult cases to determine whether or not the skin lesions will spontaneously resolve. Thirty percent of patients with LyP eventually develop lymphoma. Cabanillas suggest that the cumulative risk of developing lymphoma after 15 years approaches 80%. Hodgkin’s disease and MF are the most common malignancies associated with LyP, and Davis demonstrated the same clonal origin in a patient that had these three conditions. Treatment for LyP is cosmetic, and transiently effective but does not alter the natural course of the disease. PUVA is commonly used to treat LyP.
Actinic reticuloid is completely reversible and does not transform into lymphoma. In a recent study, Bakels showed that gene rearrangement analysis can be useful in differentiating actinic reticuloid (zero of 12 patients with a clonal population) from SS (19 of 21 samples with a clonal population). Light testing is useful in demonstrating photosensitivity. Immunophenotyping often shows a population of suppressor T cells, as opposed to helper T cells more commonly seen in MF/SS. UVB is given in gradually increasing doses to induce tolerance to light and is an effective treatment for actinic reticuloid.
Two patterns have been described histologically. One pattern is a subepidermal band-like infiltrate that resembles MF and consists of lymphocytes with medium-sized cerebriform nuclei. Epidermotropism and Pautrier’s microabscesses are rare in this histologic pattern of pseudolymphoma and may be helpful in differentiating these two processes. Medium-sized blast cells can also be seen in pseudolymphoma.
Pseudolymphoma can also mimic pleomorphic CTCL with a nodular pattern. This pattern is characterized by lymphocytes with medium- to large-sized cerebriform nuclei mixed with large blast-like lymphocytes. There are frequently large numbers of histiocytes with occasional granuloma formation seen in this pattern of pseudolymphoma.
Most of the lymphocytic infiltrate in pseudolymphomas consists of CD4+ cells, with a smaller number of CD8+ cells. The only definitive way to distinguish pseudolymphoma from CTCL is to demonstrate disappearance of the lesion with cessation of the offending drug. Bakels recently demonstrated that gene rearrangement studies and immunophenotyping may be useful in distinguishing between these two entities. None of the 10 pseudolymphoma lesions they examined demonstrated clonality or aberrant expression of T-cell antigens, whereas clonality was demonstrated in 88% and aberrant phenotype was seen in 71% of CTCLs. Recently, Brady used PCR to prove clonality in two out of 14 patients with lymphomatoid hypersensitivity. Bakels also demonstrated that the presence of CD8+ T cells, B cells, and macrophages were a characteristic finding in pseudolymphoma. There have been rare cases of hydantoin-induced pseudolymphoma that have progressed to a true cutaneous lymphoma.125 In cases of drug-induced pseudo–T-cell lymphoma, the skin lesions usually resolve completely within 4 to 6 weeks of stopping the offending drug.
T-Cell Leukemia/Lymphoma (ATLL)
The clinical presentation is variable but usually includes lymphadenopathy, hepatosplenomegaly, cutaneous lesions, and/or hypercalcemia. There are four clinical types described based on prognosis and clinicopathologic features; acute, lymphoma, chronic, and smoldering. Patients with the acute type have a median survival of 6.2 months, whereas patients with the chronic type have a median survival of 24 months. Most cases present with some peripheral-blood involvement. Skin lesions are present up to 60% of the time and can strongly resemble lesions of MF or SS. A classic example of the similarity of these lesions is demonstrated by the fact that HTLV-I was first isolated by Poiesz as stated previously, from a patient that was initially diagnosed with MF.
Traditionally, therapy has consisted of anthracycline-containing chemotherapy and/or alternating chemotherapy regimens. This approach yields a complete response rate of 17% to 42%. Standard chemotherapy does not significantly affect survival. More recently, the combination of IFN and zidovudine has been shown to be effective, attaining a major response rate of 66% in 29 patients.