Malignant T cells have a T_CM phenotype in L-CTCL and a T_EM phenotype in MF

Clonal malignant T cells can be identified in some CTCL patients by staining with commercially available antibodies to TCR Vβ subfamilies. By identifying the malignant T cell clone, researchers can assess disease burden and monitor for recurrence (7). As reported previously, clonal malignant T cells from both the blood and skin of L-CTCL patients co-expressed L-selectin and CCR7, a phenotype characteristic of T_CM (6)(Fig. 1A). Greater than 90% of malignant T cells in blood expressed CCR4, but separate populations of CLA⁻ and CLA⁺ clonal T cells existed in the blood of most patients. However, malignant T cells expressing CLA were the predominant population observed in lesional skin (Fig. 1A, Table S1).

A high forward/side scatter phenotype by flow cytometry analysis can be used to identify the malignant T cell population in MF (8). The high scatter T cell population from MF skin lesions expressed CLA and CCR4 but lacked CCR7 and L-selectin co-expression, a phenotype consistent with T_EM (Fig. 1B)(6).

A population of skin tropic T_CM are present in human skin and blood

Although T_CM were first described as L-selectin⁺/CCR7⁺ T cells that lacked expression of tissue specific adhesion receptors, our findings in CTCL suggested that a population of CLA⁺ skin tropic T_CM may exist in blood that can enter skin (9). Indeed, analysis of CD45RO⁺ memory T_CM from normal human blood demonstrated that a mean 40% (st. dev. 6.4) of circulating CD45RO⁺/L-selectin⁺/CCR7⁺ T_CM also expressed the skin homing addressin CLA (Fig 1C). In normal human skin, L-selectin/CCR7⁺ T_CM comprised approximately 20% of total T cells (Fig. 1D)(10). Similar to our observations in L-CTCL, the T_CM present in normal human skin expressed CLA, suggesting that only CLA⁺ T_CM are capable of migrating into the skin under non-inflamed conditions.

Low dose alemtuzumab is a highly effective, well tolerated therapy for L-CTCL

Alemtuzumab (10 mg) was administered subcutaneously to patients with L-CTCL three times a week for six weeks. The duration of therapy was determined by patient response, and therapy was discontinued when the skin was clinically clear. Patients were maintained on valacyclovir and trimethoprim/sulfamethoxazole for HSV and Pneumocystis jirovecii prophylaxis until six months after completion of therapy. Characteristics and clinical responses of 18 patients are shown in Table I. All patients had had CTCL for greater than 3 years with consistent skin biopsies, elevated absolute CD4 counts, and CD4/CD8 ratios >10 and were refractory to at least two prior therapies. Responding patients treated with alemtuzumab experienced rapid depletion of lymphocytes from blood and gradual clearing of skin disease. Clearance of skin disease lagged behind depletion of T cells from blood and required at least three weeks of additional therapy after loss of detectable circulating T cells (Table I). Peripheral blood disease improved in 100% of patients and completely cleared in 89%. Skin disease showed a partial response in 89% of patients and a complete response in 50%. With respect to clearing of both blood and skin disease, we observed partial responses in 89% of patients and complete remissions in 50% of patients. Responses were often durable. Patient 279 experienced complete remission for 18 months after completing a single six-week course of αCD52 (Fig. 1E) and patient 119 had a 16 month full remission after a similar course. In contrast to its effectiveness in L-CTCL, we found alemtuzumab therapy to be completely ineffective in the treatment of MF (Fig. 1F).

In general, patients treated with alemtuzumab did not develop serious infections, despite the fact that they had no circulating T or B cells (Table I). The one observed exception was a patient who had been treated with lung irradiation for a prior diagnosis of Hodgkin’s disease. This patient had multiple medical problems including radiation-induced lung fibrosis and recurrent pulmonary infections. He succumbed to respiratory distress two months after starting alemtuzumab therapy, probably secondary to pneumonia. We saw no evidence for reactivation of CMV in any of our patients treated with low dose alemtuzumab.

Alemtuzumab therapy selectively depletes the skin of malignant and benign T_CM

To gain insight into why alemtuzumab is effective in L-CTCL but not in MF, we studied T cells from the skin of L-CTCL patients before and after treatment with alemtuzumab. In L-CTCL patients in whom clonal malignant T cells could be identified, we observed depletion of the malignant T cell clone from the skin of patients after treatment with alemtuzumab; this depletion of clonal T cells from the skin correlated with clearance of skin disease and cutaneous symptoms (Table I and Fig. 2A). Further studies revealed that, in addition to loss of the malignant T cell clone, treatment with alemtuzumab depleted T_CM from skin (Fig. 2B).

Alemtuzumab therapy leaves intact a diverse population of skin resident T_EM

Skin biopsies from L-CTCL patients after alemtuzumab therapy revealed the presence of viable T cells in skin despite a complete absence of circulating T cells in these patients (Fig. 2A,B and Table S2). The surviving T cells were CD45RO⁺ memory T cells, either CD4⁺ or CD8⁺, and co-expressed the skin homing addressins CLA and CCR4 (Fig 2C). The majority of these cells were CCR7⁻/L-selectin⁻/CD27^lo, consistent with a T_EM phenotype, although a subpopulation of FOXP3⁺ regulatory T cells was also evident (Fig. 2B,C, Table S2). In general, the phenotype of T cells remaining in the skin after alemtuzumab therapy resembled T cells found in healthy skin (10). When TCR diversity was assayed by flow cytometry after staining with TCR Vβ antibodies, the TCR diversity of T cells present in skin after alemtuzumab therapy was found to be comparable to the diversity of T cells from the skin of healthy individuals (Fig. 2D).

Neutrophils participate in alemtuzumab-mediated T cell depletion

Our findings suggested that alemtuzumab depletes T cells in the circulation but not in the skin. Sparing of skin resident T cells could result from poor skin penetration of alemtuzumab, a lack of CD52 expression by skin resident T_EM, differential post-translational modification of CD52 on skin T_EM that abrogated binding, or a cell type critical for T cell depletion that is not present in the skin. We analyzed the extracellular fluid of skin samples from CTCL patients and detected the drug in the skin of patients on alemtuzumab but not in patients on other therapies (Fig. 3A), confirming that alemtuzumab enters the skin. CD52 was uniformly expressed by skin T cells from both MF and L-CTCL patients, both before and after alemtuzumab therapy (Fig 3B). We conjugated alemtuzumab to the fluorochrome Alexa Fluor 488 (AF488) and used this reagent to directly stain the T cells from the skin of both normal and CTCL patients. Alemtuzumab bound well to skin resident T cells from healthy individuals, and to both the malignant and benign T cells from CTCL skin lesions, demonstrating there were no post-translational modifications that blocked binding (Fig. 3C,D). In two mouse models, alemtuzumab has been shown to deplete CD52⁺ cells via antibody dependent cell mediated cytotoxicity (ADCC) mediated by neutrophils and to a lesser extent NK cells (11, 12). Therefore, we engrafted NOD/SCID/IL2 receptor γ chain^null mice with human T cells and treated them with alemtuzumab in the presence or absence of the neutrophil depleting Ab αLy-6G. Human T cell depletion occurred in mice treated with control antibody but not in neutrophil-depleted mice, suggesting neutrophils contribute to T cell depletion (Fig. 3E–G).

Alemtuzumab clears central memory (T_CM) but not effector memory (T_EM) disease

After antigen stimulation, T_CM normally undergo proliferation and differentiation into tissue tropic T_EM (13). In a subset of patients, clonal malignant T cells with both T_CM and T_EM phenotypes were observable in the circulation. In these patients, alemtuzumab therapy tended to improve pruritus and diffuse cutaneous erythema, clinical features of L-CTCL, but patients often had persistent or emergent fixed lesions more suggestive of MF. In one example, patient 292 had highly refractory L-CTCL and cutaneous disease that included both diffuse erythema and smaller discrete inflammatory skin lesions (Fig 4A). Malignant T cells expressed TCR Vβ17 and consisted of a mixed population of L-selectin⁺/CCR7⁺ T_CM-like cells and non-T_CM L-selectin⁻/CCR7^−/+ cells (Fig. 4B). The patient’s cutaneous symptoms improved markedly after alemtuzumab therapy but skin biopsy after treatment demonstrated residual Vβ17⁺ T cells with a T_EM phenotype. The patient underwent allogeneic stem cell transplantation (SCT) and subsequently recurred with fixed inflammatory skin lesions characteristic of MF (Fig. 4C). Topical therapy with the TLR7 agonist imiquimod, an effective treatment for MF, caused an initial central clearing and gradual resolution of these fixed lesions (Fig. 4D) (14). He remains clear of leukemic disease. Similarly, patient 414 presented with diffuse erythema, leonine facies (a clinical presentation of L-CTCL characterized by dense infiltration of the skin with malignant T cells), and more well defined nodular lesions (Fig. 4G). Analysis of T cells from the blood and skin prior to alemtuzumab therapy demonstrated the presence of both T_CM and T_EM-like subpopulations within the Vβ13.1-expressing malignant T cell pool (Fig. 4E,F). After alemtuzumab therapy, there was loss of the malignant T cell clone from blood as well as clearing of the patient’s diffuse erythema and leonine facies. However, the more discrete inflammatory nodules persisted and then worsened, requiring the addition of skin directed electron beam therapy (Fig. 4G). After electron beam treatment of his fixed lesions, the patient went into complete remission.

Skin T cells have decreased Th2 and increased Th1 responses after alemtuzumab therapy

Patients with L-CTCL have clinical abnormalities characteristic of Th2-biased immunity, including decreased antigen-specific T cell responses, impaired cell-mediated cytotoxicity, peripheral eosinophilia, and elevated levels of serum IgE and IgA (15–17). Malignant T cells in L-CTCL have been shown to be Th2 biased (18). One prior study demonstrated improved Th1 responses in peripheral blood T cells after the immunomodulatory therapies extracorporeal photopheresis and/or IFNα therapy (19). We assayed cytokine production of skin T cells before and after alemtuzumab (Fig. 5). IL-4 was produced by greater percentages of T cells in L-CTCL lesions than in normal skin (Fig. 5A). Alemtuzumab therapy reduced IL-4 production to levels observed in normal skin. Production of IFNγ was slightly higher in lesional L-CTCL skin than in normal controls, but alemtuzumab therapy further enhanced IFNγ to levels significantly higher than those observed in normal skin. In contrast, IL-17 production in both untreated and alemtuzumab-treated L-CTCL patients was significantly lower than in normal skin. TNFα is produced by most T cells in healthy skin and although alemtuzumab enhanced TNFα production, increases were not statistically significant. TSLP levels were not increased in the lesional skin of patients with MF or L-CTCL (Fig. S1).

Down-regulation of CD52 and development of a blocking antibody can lead to alemtuzumab resistance

Two patients developed resistance to alemtuzumab therapy. Patient 295 achieved two complete remissions but only a partial response to a third course. T cells were not fully depleted during this third course and three weeks after beginning therapy, he had both circulating CD4⁺ and CD8⁺ T cells (Fig. 6A). Malignant T cells were not recognized by available TCR Vβ antibodies but could be identified by their CD4⁺/CD3^low phenotype. Peripheral blood disease worsened after discontinuation of alemtuzumab and analysis demonstrated loss of CD52 expression on malignant T cells and a subset of CD8 cells (Fig. 6B). Loss of CD52 expression was durable; six months after alemtuzumab therapy, lymphocytes remained CD52⁻ (Fig. 6C).

Patient 042 experienced an initial drop and then a rebound in the number of circulating monocytes and T cells on alemtuzumab therapy (Fig. 6D). After six weeks, normal numbers of T cells and monocytes expressing CD52 were demonstrable in the peripheral blood (Fig. 6E). To assay for the presence of an inhibitory factor in serum, AF488-conjugated alemtuzumab was used to stain normal blood T cells in the presence and absence of plasma from patient 042 or healthy donors. Plasma from patient 042 blocked binding of alemtuzumab to T cells (Fig. 6F). Inhibition of binding was titratable (Fig. 6G). Preclearance of patient 042’s plasma with protein G-conjugated sepharose beads led to a loss of binding inhibition, suggesting the presence in this patient plasma of an antibody that blocked alemtuzumab binding to CD52.

Skin and blood samples

The protocols of this study were performed in accordance with the Declaration of Helsinki, and were approved by the Institutional Review Board of the Partners Human Research Committee (Partners Research Management). Skin from healthy patients was obtained from patients undergoing cosmetic surgery procedures and blood from healthy individuals was obtained as discarded tissue after leukopheresis. Blood and lesional skin from patients with CTCL were obtained from patients seen at the Dana-Farber/Brigham and Women’s Cancer Center Cutaneous Lymphoma Program. L-CTCL and MF patients described in this manuscript met the WHO-EORTC criteria for L-CTCL/SS or MF (1). Since 2002, 443 CTCL patients have been enrolled in studies that permit the collection of blood and/or skin; of these, 427 patients (96%) have consented to provide samples. There have been 11 publications to date on this patient population. Patients are assigned a unique study number (e.g. Pt 188) that is used to identify their experimental data in this and other manuscripts arising from studies of this patient population. PBMC were isolated by ficoll centrifugation and T cells were isolated from skin using short-term explant cultures (1–3 weeks) as described (49).

Flow cytometry

Analysis of T cells was performed using directly conjugated monoclonal antibodies obtained from BD Biosciences, eBioscience, Biolegend, or R&D Systems. Vβ staining was performed using the IOtest Beta Mark TCR V beta Repertoire kit (Beckman Coulter) as per manufacturer’s instructions. FOXP3 staining utilized the PCH101 antibody (eBioscience). Isotype-matched negative control antibodies were used to set the gates for positive staining. Alemtuzumab was conjugated to AF488 using Alexa Fluor 488 Microscale Protein Labeling Kit (Invitrogen) as per manufacturer’s instructions. For analysis of cytokine production, T cells were stimulated with either control medium or 50 ng/ml PMA and 750 ng/ml ionomycin plus 10 μg/ml Brefeldin A (Calbiochem) for four hours. Cells were surface stained, fixed, permeabilized, stained with anti-cytokine antibodies, and examined by flow cytometry. Analysis of flow cytometry samples was performed on Becton Dickinson FACSCanto instruments and data was analyzed using FACSDiva software (V5.1).

Detection of alemtuzumab in the extracellular fluid of skin

Alemtuzumab is a humanized monoclonal antibody that retains amino acid sequences derived from the original rat IgG2a antibody (50). These residual rat sequences can be recognized by polyclonal anti-rat antibodies, allowing the detection of alemtuzumab in biological specimens (51). Briefly, skin biopsies from patients were minced in 100 μl of medium and the resulting supernatant was denatured in the presence of 2% SDS for 5 min at 96°C to expose the rat-derived amino acid sequence present in alemtuzumab. This denatured supernatant was then incubated with 6.9 μM carboxylated polystyrene beads (Bangs Laboratories) coated with rabbit anti-rat IgG antibodies (Sigma Aldrich). PE-labeled rabbit anti-human IgG (BioLegend) was then added to the beads and binding was assessed by flow cytometry. 0.01 ug/ml alemtuzumab in PBS was used as a positive control (Fig. 3A, far right).

T-cell depletion studies in immunodeficient mice engrafted with human T cells

T cells from the blood of healthy donors were obtained using the Pan T Cell Isolation Kit (Miltenyi Biotec) and the autoMACS^™ Separator (Miltenyi Biotec) according to manufacturer’s instructions. 6- to 8-week-old NOD/SCID/IL2 receptor γ chain^null mice were injected I.V. with 2.5×10⁶ human T cells in 250 μL of saline. Successfully engrafted mice were those that had greater than 500/μL human CD3⁺ cells in blood as determined by flow cytometry staining for CD3 and direct cell counting. Mice were then injected I.P. with 0.5 mg anti-mouse Ly-6G abs (1A8; BioXCell) or control rat IgG2a (BioXCell) and 72 hours later, 2.5 μg alemtuzumab or control human IgG (Jackson ImmunoResearch) was administrated I.P.. The efficacy of T cell depletion was measured by comparing the absolute numbers of CD3 positive cells/μL blood before and 24 hrs after alemtuzumab administration. The non-parametric Wilcoxon rank-sum test was used for comparisons between groups.

Depletion of blocking antibody from plasma using protein G

To deplete immunoglobulins from the plasma of patient 042, plasma was incubated with Pierce^® Protein G Agarose (Thermo Scientific, Fisher) at 4 °C overnight. Ig-depleted and non-depleted plasma from patient 042 and plasma from healthy donors were then assessed for the ability to block alemtuzumab binding. Plasma samples were mixed at the indicated concentrations with Alexa Fluor 488-labeled alemtuzumab and this mixture was used to label PBMC derived from the blood from healthy individuals. Binding of fluorochrome-labeled alemtuzumab to cells was then assessed by flow cytometry.

Statistical analyses

The non-parametric Wilcoxon rank-sum test was used for comparisons between two groups. Statistical analysis of three comparison groups was performed using the Kruskal-Wallis test and Dunn’s test. A significance level of P<0.05 was chosen for all analyses.