Primary cutaneous T-cell lymphomas (CTCL) are rare malignancies characterized by clonal T-cell infiltrates (Kim et al, 2005; Prince et al, 2009). Mycosis fungoides (MF), the most common form of CTCL, presents with patches, plaques, tumours and erythroderma whereas Sézary syndrome (SS) is characterized by erythroderma, lymphadenopathy, and Sézary cells in the peripheral blood. Although CTCL often pursues an indolent course, a poor prognosis has been reported in patients with tumoural/ulcerated lesions, erythroderma, and transformation into T-cell lymphoma (Kim et al, 2005; Prince et al, 2009). Advanced stages of CTCL often show resistance to available treatment (Kim et al, 2005; Cheson, 2007; Prince et al, 2009). Two unconjugated humanized monoclonal antibodies (MAbs) have been used recently in the treatment of CTCL. Alemtuzumab, a CD52 MAb, has proved a valuable treatment option for several subtypes of T-cell malignancy. In CTCL, clinical response rates of 84% (47% complete remission, CR) have been demonstrated in SS patients (Ravandi & O’Brien, 2005; Bernengo et al, 2007). Long-term, durable responses may be obtained with alemtuzumab (DBJK and MJSD, unpublished observations). Zanolimumab is a CD4 MAb that has shown activity in patients with CTCL and non-CTCL lymphomas (d’Amore et al, 2010). In two Phase II studies, 56% of MF patients responded to high dose of zanolimumab whilst for SS patients the response rates were 25% (Kim et al, 2007; d’Amore et al, 2010). Here, we present a case of SS who, whilst initially sensitive to alemtuzumab, became alemtuzumab-resistant due to selective loss of CD52 expression. Nevertheless, it was possible to induce a second durable remission with zanolimumab. A 68-year-old Caucasian woman was diagnosed with SS in 2001. She presented with ‘resistant-to-treatment eczema’ and inconclusive skin biopsies, but later developed extensive exfoliative erythema. There was a low-level lymphocytosis composed of Sézary cells expressing CD4 and CD52. Polymerase chain reaction of the T-cell receptor γ locus (TRG@) of skin, lymph node and peripheral blood samples showed identical clonal rearrangements. Full body computerized tomography (CT) scan was normal. The patient was managed with topical and systemic steroids, PUVA (psoralen and ultraviolet A) photochemotherapy, ciclosporin, methotrexate and acitretin with little effect. By 2004, she had developed erythroderma and was symptomatic with fever, sweats and shivering (Fig 1A). Skin biopsy showed a mild upper dermal inflammatory infiltrate composed of small lymphocytes with extension into the overlying epidermis with associated spongiosis. No classical Pautrier microabscesses were observed. Malignant T-cells expressed CD4 (Fig 1B). She was treated with subcutaneous alemtuzumab 30 mg initially on a daily dose schedule. Her symptoms resolved within 72 h. After the first week, doses of alemtuzumab were given three times per week for a total of 12 weeks. This therapy induced a CR that lasted for 18 months. (A) Generalized, exfoliative erythroderma seen in this patient before treatment with either alemtuzumab or zanolimumab. (B) Skin biopsy stained with CD4 showing a mild upper dermal infiltrate composed of generally small lymphocytes extending into the epidermis with associated spongiosis. No classical Pautrier microabscesses were observed (Original magnification ×20). At relapse, there were no Sézary cells in the peripheral blood. Alemtuzumab was reinstated but neither this nor intravenously administered MAb nor the addition of high dose methylprednisolone made any significant impression on her disease. Neutralizing alemtuzumab anti-idiotype antibodies were sought but none detected. Alemtuzumab was abandoned. Two months later the patient was commenced on zanolimumab. Zanolimumab (980 mg per infusion) was administered initially twice a week for 2 weeks. The patient’s symptoms and erythroderma were slow to resolve but after 4 months of treatment (17 infusions), she had achieved almost CR: the patient was well and had only mild erythema on thighs and torso. Analysis of peripheral blood identified no detectable CD4 lymphocytes. Zanolimumab was discontinued and CD4 lymphopenia resolved slowly over a period of 12 months without any infective complications. One month after stopping zanolimumab the patient suddenly developed macular, blanching, urticarial patches affecting 80% of body and became systemically unwell (Fig 2). Investigations showed normal C-reactive protein and full blood count. Skin cultures yielded Staphylococcus aureus and beta-hemolytic Streptococcus group B. Skin biopsy exhibited non-specific findings and showed no histopathological, immunocytochemical or molecular features of CTCL. The patient received erythromycin and mupirocin with resolution of signs and symptoms within 2 weeks. She remained well until she relapsed with erythroderma and peripheral blood Sézary cells 12 months later. At the time of relapse, peripheral blood Sézary cells expressed CD4 but not CD52, suggesting that the lack of efficacy of alemtuzumab on the second treatment was due to selection of cells lacking expression of CD52. Labile urticarial eruption affecting 80% of the body surface following zanolimumab. Zanolimumab blocks T-cell activation via CD4 and also facilitates destruction of target cells via antibody-dependent cellular cytotoxicity (ADCC) (Rider et al, 2007); which of these two possible mechanisms is of pivotal importance in the treatment of T-cell malignancy is not yet clear. In comparison with alemtuzumab, where therapeutic responses are (as in our patient) often very rapid, responses to zanolimumab are usually slow with a median time to response of between 2 and 12 weeks) (Kim et al, 2007). These data may indicate that the major mode of action of this MAb is through blocking of T-cell activation rather than ADCC. This is supported by the observation that CD4 internalises after zanolimumab cross-linking, rendering cells effectively CD4 negative. In contrast, alemtuzumab does not internalize significantly and remains available to mediate ADCC even in vivo (Dyer et al, 1989). It is interesting to note that in this case the CD4 expression was maintained after zanolimumab therapy, whilst CD52 expression was undetectable. The optimal doses of zanolimumab, the role of zanolimumab in combination with other agents and, in particular, the possible role of maintenance zanolimumab therapy in responding patients remains to be determined. Despite the complete absence of CD4 cells in the peripheral blood that follows its use, the MAb does not appear to be associated with a significant increase in opportunistic infections, which may allow for more protracted courses of this MAb to be given in order to improve and maintain responses. Our patient experienced an apparent ‘flare’ of disease but this was due to Staphylococcus aureus infection. Exacerbation of CTCL triggered by Staphylococcus aureus has been reported previously. It should also be noted that eczematous dermatitis has also been observed in a proportion of patients receiving zanolimumab and appears to correlate with response. We thank Genmab (Copenhagen DK) for generously providing zanolimumab for our patient on compassionate request basis; Genmab has licensed zanolimumab to TenX Biopharma Inc (Philadelphia, PA) for further development and commercialization. We thank Professor Geoff Hale (BioAnalab, Oxford UK) for performing anti-globulin assays.