Prognostic factors, prognostic indices and staging in mycosis fungoides and Sézary syndrome: where are we now?

Mycosis fungoides is the most prevalent form of primary cutaneous T-cell lymphoma. Patients frequently present with early-stage disease typically associated with a favourable prognosis and survival of 10-35 years, but over 25% may progress to advanced disease with a median survival < 4 years, and just 13 months in those with nodal involvement. Sézary syndrome presents in advanced disease with erythroderma, blood involvement and lymphadenopathy. The Bunn and Lamberg staging system (1979) includes stages IA-IIA (early-stage disease) and IIB-IVB (advanced-stage disease) and provides prognostic information, but some patients with tumour-stage disease (IIB) have a worse prognosis than those with erythrodermic-stage (III). Conversely, patients with plaque-stage (IB) folliculotropic mycosis fungoides may have a worse outcome than those with tumour-stage (IIB). The more recent staging system of the European Organisation for the Research and Treatment of Cancer/International Society for Cutaneous Lymphoma has been designed to reflect tumour burden at different sites. However, this staging system has not been validated prospectively for prognosis. Furthermore, this staging system does not include a detailed measurement of skin tumour burden, as indicated by the modified skin weighted severity assessment tool. This assessment measures body surface area of disease and is weighted to record patch, plaque and tumour to produce a numerical value from 0·5 to 400 and is an established endpoint for clinical studies. Nor does this staging include clinicopathological features associated with a poor prognosis such as folliculotropism. Here we review the clinical, haematological, pathological and genotypic parameters outside the staging system, which may affect survival in mycosis fungoides and Sézary syndrome. Most studies are retrospective and single centre. The identification of poor prognostic factors may be used to develop a prognostic index to use alongside staging, which may be of benefit in mycosis fungoides/Sézary syndrome to identify patients with a potentially poor prognosis.

A cutaneous lymphoma international prognostic index (CLIPi) for mycosis fungoides and Sezary syndrome.

BACKGROUND: There is no prognostic index for primary cutaneous T-cell lymphomas such as mycosis fungoides (MF) and Sezary syndrome (SS). METHOD: Two prognostic indices were developed for early (IA-IIA) and late stage (IIB-IVB) disease based on multivariate data from 1502 patients. End-points included overall survival (OS) and progression free survival (PFS). External validation included 1221 patients. FINDINGS: Significant adverse prognostic factors at diagnosis consisted of male gender, age >60, plaques, folliculotropic disease and stage N1/Nx for early stage, and male gender, age >60, stages B1/B2, N2/3 and visceral involvement for late stage disease. Using these variables we constructed two separate models each defined using 3 distinct groups for early and late stage patients: 0-1 (low risk), 2 (intermediate risk), and 3-5 factors (high risk). 10 year OS in the early stage model was 90.3% (low), 76.2% (intermediate) and 48.9% (high) and for the late stage model 53.2% (low), 19.8% (intermediate) and 15.0% (high). For the validation set significant differences in OS and PFS in early stage patients (both p<0.001) were also noted. In late stage patients, only OS differed between the groups (p=0.002). INTERPRETATION: This proposed cutaneous lymphoma prognostic index provides a model for prediction of OS in early and late stage MF/SS enabling rational therapeutic choices and patient stratification in clinical trials.

Promoter methylation of argininosuccinate synthetase-1 sensitises lymphomas to arginine deiminase treatment, autophagy and caspase-dependent apoptosis.

Tumours lacking argininosuccinate synthetase-1 (ASS1) are auxotrophic for arginine and sensitive to amino-acid deprivation. Here, we investigated the role of ASS1 as a biomarker of response to the arginine-lowering agent, pegylated arginine deiminase (ADI-PEG20), in lymphoid malignancies. Although ASS1 protein was largely undetectable in normal and malignant lymphoid tissues, frequent hypermethylation of the ASS1 promoter was observed specifically in the latter. A good correlation was observed between ASS1 methylation, low ASS1 mRNA, absence of ASS1 protein expression and sensitivity to ADI-PEG20 in malignant lymphoid cell lines. We confirmed that the demethylating agent 5-Aza-dC reactivated ASS1 expression and rescued lymphoma cell lines from ADI-PEG20 cytotoxicity. ASS1-methylated cell lines exhibited autophagy and caspase-dependent apoptosis following treatment with ADI-PEG20. In addition, the autophagy inhibitor chloroquine triggered an accumulation of light chain 3-II protein and potentiated the apoptotic effect of ADI-PEG20 in malignant lymphoid cells and patient-derived tumour cells. Finally, a patient with an ASS1-methylated cutaneous T-cell lymphoma responded to compassionate-use ADI-PEG20. In summary, ASS1 promoter methylation contributes to arginine auxotrophy and represents a novel biomarker for evaluating the efficacy of arginine deprivation in patients with lymphoma.

Constitutive activation of STAT3 in Sézary syndrome is independent of SHP-1.

Constitutive and persistent activation of STAT3 has been implicated in the pathogenesis of many malignancies. Studies of CTCL cell lines have previously suggested that aberrant activation of STAT3 is mediated via silencing of the negative regulator SHP-1 by promoter methylation. In this study of ex vivo tumour cell populations from 18 Sézary syndrome (SS) patients, constitutive phosphorylation of STAT3, JAK1 and JAK2 was present in all patients, but was absent in comparative CD4+ T-cells from healthy controls. Furthermore, no loss or significant difference in SHP-1 expression was observed between patients and healthy control samples. Methylation-specific PCR analysis of the SHP-1 CpG island in 47 SS patients and 11 healthy controls did not detect any evidence of methylation. Moreover, small interfering RNA knockdown of SHP-1 had no effect on phosphorylation of STAT3. In contrast, treatment of SS tumour cells with the pan-JAK inhibitor pyridone 6 led to downregulation of phosphorylated STAT3 (pSTAT3), its target genes and induction of apoptosis. No evidence for common JAK1/JAK2-activating mutations was found. These data demonstrate that constitutive activation of STAT3 in SS is not due to the loss of SHP-1, but is mediated by constitutive aberrant activation of JAK family members.

Bexarotene therapy for mycosis fungoides and Sézary syndrome.

BACKGROUND: Bexarotene (Targretin) is a synthetic retinoid which is licensed for the treatment of advanced refractory cutaneous T-cell lymphoma (CTCL). OBJECTIVES: To summarize our experience with bexarotene for patients with CTCL with the aim of assessing efficacy and safety. METHODS: A retrospective study of 66 patients (44 male, 22 female) with mycosis fungoides (40 patients) or Sézary syndrome (26 patients) who were commenced on bexarotene prior to August 2007 was carried out. Nineteen patients had early-stage (IB-IIA) refractory mycosis fungoides and 47 patients had advanced-stage CTCL (IIB-IVB). RESULTS: Fifty-two out of 66 (79%) patients completed over 1 month of therapy with an intention-to-treat response rate of 44% (29/66). Of the patients, six (9%) had a complete response, 23 (35%) had a partial response, 15 (23%) had stable disease and eight (12%) had progressive disease. Median time to maximal response was 3 months (1-9 months). Median response duration was 8 months (1 to > 48 months). Median time to progression was 9 months (3-44 months). Fourteen patients (21%) did not complete a month of bexarotene therapy. Adverse effects of the whole group included central hypothyroidism in 100% (all grade II and managed with thyroid replacement) and hyperlipidaemia in 100% (all managed with lipid-lowering therapy +/- dose reduction). Responses were seen in all stages and were higher in advanced stages: 26% (five of 19) with early-stage and 51% (24/47) of advanced-stage disease. Responses were seen in skin, blood and lymph nodes. Twenty-eight out of 66 patients were treated with bexarotene monotherapy and the remainder were on one or more additional anti-CTCL therapies. CONCLUSIONS: Our data demonstrate that bexarotene is well tolerated in most patients and responses are seen in almost half of patients with all disease stages. However partial responses were not graded and would include any improvement seen in the skin, blood and lymph node.

Folliculotropic mycosis fungoides (stage IIA) progressing to Sézary syndrome: a case report.

Folliculotropic mycosis fungoides is associated with a worse prognosis than classical mycosis fungoides (MF), but whether this is due to resistance to skin-directed therapy or to biological differences is unclear. We discuss a case of a patient with folliculotropic MF (stage IIA) who progressed to develop Sézary syndrome (SS), stage IVB, over 6 years. A 40-year-old man presented with pruritic plaques affecting his head and trunk, characterized by follicular plugging. The histology was consistent with folliculotropic MF and T-cell gene analysis studies revealed a T-cell clone in the skin only. His condition gradually deteriorated and 5 years after presentation, T-cell gene analysis studies revealed the presence of a clone in the blood identical with that seen in the skin. His condition progressed with the development of erythrodermic disease and a leukaemic blood picture and he subsequently died of systemic nodal and visceral involvement. We present the first report detailing the stepwise progression of a patient with stage IIA folliculotropic MF to SS. This case demonstrates that MF and SS represent a clinical spectrum of the same disease.

Cerebellar ataxia in sheep grazing pastures infested with Romulea rosea (onion grass or Guildford grass).

There have been anecdotal reports since 1962 of 'staggers' in sheep grazing Romulea rosea infested pastures, but this is the first detailed account. In September 2005, a locomotor disorder developed in 12 of 120 Merino wethers that had grazed R. rosea infested pasture at Albury, New South Wales, over several months. Affected sheep displayed signs that included limb paresis, knuckling over in the fetlocks, fine head tremor, incoordination, and an equilibrium disturbance characterised by frequent falling. The microscopic examination of brain and spinal cord tissues from two affected sheep revealed mild vacuolation, occasional lymphocytic cuffing around blood vessels, mild Wallerian degeneration, and occasional glial cells that contained honey-brown cytoplasmic pigments. The most significant changes were found in the cerebellum, where there were decreased numbers of Purkinje cells, increased numbers of glial cells, scattered vacuoles and occasional swollen axons. Previous reports of cerebellar toxicoses in ruminants have involved goats and cattle and have been associated with the ingestion of six Solanum spp. The Purkinje cell loss in this type of disorder is ultimately extensive and consequently affected animals may survive, but will remain permanently disabled.

An extended outbreak of congenital chondrodysplasia in calves in South East Australia.

OBJECTIVE: To report an outbreak of congenital chondrodystrophy in calves in South East Australia. METHODS: District veterinarians investigated reported cases of calf deformities. Owners of affected herds were interviewed using a standard questionnaire to identify potential risk factors. Dams of several affected calves were serologically tested for Akabane virus, Aino virus, pestivirus and bluetongue, and affected calves were tested for pestivirus antigen and serum immunoglobulin concentrations. Gross and histopathological examinations of numerous calves were performed, concentrating on the musculoskeletal system. RESULTS: A case definition of distinctive skeletal deformities was established, and 89 property owners reported calves with chondrodystrophy in Spring 2003, 2004 or 2005. Some 14 property owners reported affected calves in more than one year. Prevalence and severity of deformity varied greatly between and within properties. None of breed, sex, age of dam, lineage, pasture type, supplementary feeding, fertiliser use or toxic plants was consistently associated with the disease. All dams experienced hot, dry conditions during the first trimester of pregnancy and were exposed to adverse conditions thereafter. Consistently dams were reported to have been grazing undulating to hilly terrain during early pregnancy. All serological tests were negative for Akabane virus, Aino virus, pestivirus and bluetongue. Histopathology of affected skeletal samples showed chondrodysplasia. CONCLUSION: The outbreak had similarities with previous outbreaks reported in the region. No specific aetiology could be determined. There is some evidence that the cause of the deformities could be a manganese deficiency during foetal development. Ongoing work to test this hypothesis is therefore warranted.

Recurrent finding of the FIP1L1-PDGFRA fusion gene in eosinophilia-associated acute myeloid leukemia and lymphoblastic T-cell lymphoma.

The FIP1L1-PDGFRA fusion gene has been described in patients with eosinophilia-associated myeloproliferative disorders (Eos-MPD). Here, we report on seven FIP1L1-PDGFRA-positive patients who presented with acute myeloid leukemia (AML, n=5) or lymphoblastic T-cell non-Hodgkin-lymphoma (n=2) in conjunction with AML or Eos-MPD. All patients were male, the median age was 58 years (range, 40-66). AML patients were negative for common mutations of FLT3, NRAS, NPM1, KIT, MLL and JAK2; one patient revealed a splice mutation of RUNX1 exon 7. Patients were treated with imatinib (100 mg, n=5; 400 mg, n=2) either as monotherapy (n=2), as maintenance treatment after intensive chemotherapy (n=3) or in overt relapse 43 and 72 months, respectively, after primary diagnosis and treatment of FIP1L1-PDGFRA-positive disease (n=2). All patients are alive, disease-free and in complete hematologic and complete molecular remission after a median time of 20 months (range, 9-36) on imatinib. The median time to achievement of complete molecular remission was 6 months (range, 1-14). We conclude that all eosinophilia-associated hematological malignancies should be screened for the presence of the FIP1L1-PDGFRA fusion gene as they are excellent candidates for treatment with tyrosine kinase inhibitors even if they present with an aggressive phenotype such as AML.

Molecular staging of lymph nodes from 60 patients with mycosis fungoides and Sézary syndrome: correlation with histopathology and outcome suggests prognostic relevance in mycosis fungoides.

BACKGROUND: Histological evidence of lymph node involvement is associated with a poor prognosis in patients with cutaneous T-cell lymphoma (CTCL). OBJECTIVES: To determine whether T-cell receptor (TCR) gene analysis is of prognostic relevance in CTCL. METHODS: TCR gene analysis was performed on lymph node specimens from 60 patients with mycosis fungoides (MF) and Sézary syndrome (SS) using a highly sensitive polymerase chain reaction (PCR)/single-strand conformational polymorphism analysis and results were correlated with skin, overall clinical and histological lymph node stages. RESULTS: The frequency with which a T-cell clone was detected in lymph node samples from patients with MF increased with skin stage, overall clinical stage and with the degree of histological involvement: six of 19 patients with uninvolved lymph nodes or limited histological involvement (LN0-2) and 13 of 14 patients with advanced histological involvement (LN3-4) had a detectable T-cell clone. In SS, 22 of 27 patients had a detectable lymph node T-cell clone. The clonal patients had a poorer prognosis than nonclonal patients (median survival from biopsy of > 72 months vs. 16 months for MF and 41.5 vs. 16.5 months for SS). Regression analysis confirmed that TCR gene analysis identifies a group of MF patients with a worse prognosis (P = 0.013). However, the molecular lymph node stage did not provide independent prognostic information in this cohort of patients in multivariate analysis. CONCLUSIONS: Molecular staging in MF and SS using a PCR-based method for TCR gene analysis provides additional information to histological examination. Specifically, this study identified a group of MF patients with early lymph node involvement with a poorer prognosis. However, a larger prospective study of patients with MF and early histological lymph node involvement is required to confirm whether molecular staging of lymph nodes provides independent prognostic information in a multivariate model.

Lymphomatoid papulosis in a patient with atopic eczema on long-term ciclosporin therapy.

We report a 36-year-old man with atopic eczema who developed lymphomatoid papulosis while taking ciclosporin. Latent membrane protein 1 and in situ hybridization for Epstein-Barr virus were negative. There are only two reports in the literature of patients taking ciclosporin to control atopic eczema who developed primary cutaneous CD30+ T-cell lymphoproliferative disorders. The development of T-cell lymphoproliferative disorders including lymphomas is well described in patients with solid organ transplants who are taking ciclosporin. Also, it has been noted in patients taking ciclosporin for rheumatological conditions or psoriasis.

BCL2 and JUNB abnormalities in primary cutaneous lymphomas.

BACKGROUND: BCL2 is upregulated in nodal and extranodal B-cell non-Hodgkin's lymphomas, with a consequent antiapoptotic effect. However, loss of BCL2 has also been noted in some malignancies, suggesting a different molecular pathogenesis. OBJECTIVES: To investigate genomic and protein expression status of BCL2 and to compare the results with that of JUNB in primary cutaneous lymphomas (PCLs). METHODS: We analysed gene copy number of BCL2 and JUNB in 88 DNA samples from 80 patients with PCL consisting of Sézary syndrome/mycosis fungoides (SS/MF), primary cutaneous B-cell lymphoma (PCBCL) and primary cutaneous CD30+ anaplastic large cell lymphoma (C-ALCL) by the use of real-time polymerase chain reaction (PCR) and immunohistochemistry (IHC). Real-time PCR and IHC findings were subsequently compared with the results of additional fluorescent in situ hybridization (FISH) analysis of 23 cases of SS and Affymetrix cDNA expression microarray study of two primary cutaneous T-cell lymphoma (CTCL) cell lines. RESULTS: Real-time PCR analysis showed loss of BCL2 gene copy number in 22 of 80 PCL cases (28%), including 17 of 42 SS/MF, three of 13 C-ALCL and two of 33 PCBCL samples, and gain of BCL2 in four PCBCL samples. Gain of JUNB was identified in 18 of 71 PCL cases (25%), including nine of 35 SS/MF, seven of 13 C-ALCL and two of 31 PCBCL samples. IHC analysis revealed absent nuclear expression of BCL2 protein in 47 of 73 PCL cases, comprising 28 of 36 SS/MF, eight of eight C-ALCL and 11 of 29 PCBCL cases. In contrast, BCL2 protein expression was detected in 26 of 73 PCL cases, consisting of 18 of 29 PCBCL and eight of 36 SS/MF cases. JUNB protein expression was present in tumour cells from 30 of 33 of SS/MF and eight of eight C-ALCL, and was absent in tumour cells from 18 of 27 PCBCL cases. A comparison between BCL2 and JUNB revealed loss of BCL2 and gain of JUNB in five of 35 SS/MF samples, and expression of JUNB protein and absent BCL2 expression in 25 SS/MF and eight of eight C-ALCL cases. In contrast, expression of BCL2 and absent JUNB expression were detected in 67% of PCBCL cases. Additional FISH analysis revealed deletion of BCL2 in 19 of 23 SS cases (83%), including eight cases with BCL2 loss shown by real-time PCR. Furthermore, Affymetrix expression microarray demonstrated decreased expression of proapoptotic and antiapoptotic genes involved in BCL2 signalling pathways such as BOK, BIM, HRK, RASA1 and STAT2 in two CTCL cell lines with BCL2 loss and absent BCL2 expression. Increased expression of JUNB was also identified in the MF cell line. CONCLUSIONS: These findings provide a comprehensive assessment of BCL2 and JUNB status in PCL, and suggest that there is a selection pressure in a subset of CTCL cases for tumour cells showing BCL2 loss and upregulation of JUNB primarily through chromosomal deletion and amplification, respectively.

A randomized cross-over study to compare PUVA and extracorporeal photopheresis in the treatment of plaque stage (T2) mycosis fungoides.

PUVA is a well-established and effective treatment for plaque stage mycosis fungoides (MF) but its use is limited on a long-term basis because of the risk of cutaneous carcinogenesis. A further disadvantage is that nonexposed areas (sanctuary sites) often develop persistent disease. Therefore it is important to find alternative methods of treatment. Extracorporeal photopheresis (ECP) is a form of photochemotherapy that involves exposure of white blood cells to UVA with psoralens and can be effective in Sézary syndrome and erythrodermic cutaneous T-cell lymphoma. The aim of this study was to compare the efficacy of PUVA and ECP in the treatment of patients with T2 plaque stage (Stage 1B) MF who had a detectable peripheral blood T-cell clone. The study was of a cross-over design. Sixteen patients were randomized to receive either PUVA twice weekly for 3 months followed by ECP once monthly for 6 months at relapse, or vice-versa. Response was assessed by monthly skin scores and peripheral blood T-cell clonality. Ten patients received PUVA initially and six ECP initially. Eight patients completed the study. Skin scores taken at the completion of each treatment arm in patients who completed the study were 113 units better (confidence interval, 42-184 units) following 3 months PUVA than 6 months ECP (P = 0.002). Peripheral blood T-cell clones were detectable in all patients post-treatment. This study indicates that ECP is not effective in the treatment of plaque stage (1B/T2) MF even in patients with molecular evidence of a peripheral blood T-cell clone. Although PUVA was more effective than ECP, neither treatment modality cleared malignant T-cells from the peripheral blood.

Cutaneous Waldenström's macroglobulinaemia.

We report a case of Waldenström's macroglobulinaemia that presented with infiltrated skin nodules and plaques but without systemic symptoms. Cutaneous manifestations such as purpura, oedema, urticaria and ulceration may be seen in Waldenström's macroglobulinaemia and are a consequence of hyperviscosity, cryoglobulinaemia and/or tissue deposition of immunoglobulins. Direct cutaneous infiltration by neoplastic lymphoid cells is less common, and very rare as the initial presentation of Waldenström's macroglobulinaemia.