Transient T cell depletion causes regression of melanoma metastases

Abstract

Background: Cognate immunity against neoplastic cells depends on a balance between effector T cells and regulatory T (Treg) cells. Treg cells prevent immune attack against normal and neoplastic cells by directly suppressing the activation of effector CD4+ and CD8+ T cells. We postulated that a recombinant interleukin 2/diphtheria toxin conjugate (DAB/IL2; Denileukin Diftitox; Ontak) may serve as a useful strategy to deplete Treg cells and break tolerance against neoplastic tumors in humans.

Methods: We administered DAB/IL2 (12 microg/kg; four daily doses; 21 day cycles) to 16 patients with metastatic melanoma and measured the effects on the peripheral blood concentration of several T cell subsets and on tumor burden.

Results: We found that DAB/IL2 caused a transient depletion of Treg cells as well as total CD4+ and CD8+ T cells (< 21 days). T cell repopulation coincided with the de novo appearance of melanoma antigen-specific CD8+ T cells in several patients as determined by flow cytometry using tetrameric MART-1, tyrosinase and gp100 peptide/MHC conjugates. Sixteen patients received at least one cycle of DAB/IL2 and five of these patients experienced regressions of melanoma metastases as measured by CT and/or PET imaging. One patient experienced a near complete response with the regression of several hepatic and pulmonary metastases coupled to the de novo appearance of MART-1-specific CD8+ T cells. A single metastatic tumor remained in this patient and, after surgical resection, immunohistochemical analysis revealed MART1+ melanoma cells surrounded by CD8+ T cells.

Conclusion: Taken together, these data indicate that transient depletion of T cells in cancer patients may disrupt the homeostatic control of cognate immunity and allow for the expansion of effector T cells with specificity against neoplastic cells. Several T cell depleting agents are clinically available and this study provides strong rationale for an examination of their efficacy in cancer patients.

Trial registration: NCT00299689 (ClinicalTrials.gov Identifier).

Figures

Figure 1

DAB/IL2 transiently depletes CD4+ and CD8+ T cells in melanoma patients. 10 patients with stage IV metastatic melanoma were administered DAB/IL2 (intravenous; 12 μg/kg) daily × 4 days (arrows indicate each administration). Whole blood was collected on the indicated days and analyzed for absolute lymphocyte (black), granulocyte (red) and monocyte (green) concentration with an automated hematology analyzer (A) and absolute CD4+ and CD8+ T cell concentration by flow cytometry (B). The peripheral blood concentrations of CD4+ and CD8+ T cells were quantified by multiplying the percentage of anti-CD4 or anti-CD8 fluorescence-positive cells within the lymphocyte forward/side scatter gate by the absolute lymphocyte concentration determined using an automated hematology analyzer. Percent control of each sample was calculated by dividing the absolute cell concentration on the indicated day of treatment with the absolute cell concentration on day 0 prior to DAB/IL2 administration (× 100). Data are represented as averages ± standard error of the mean (n = 10 patients).

Figure 2

DAB/IL2 transiently depletes CD4+/CD25HI/Foxp3+ T cells. Whole blood was collected from patient P9 during cycle one of DAB/IL2 administration just prior to the first (day 0) and last dose (day 3) and then 7 and 21 days after initiation of DAB/IL2 therapy. The peripheral blood mononuclear cells were isolated from the whole blood by Ficoll gradient centrifugation and stained with fluorescent conjugates of monoclonal antibodies specific for CD4, CD25 and Foxp3. In order to quantify the percentage of CD4+/CD25HI/Foxp3+ T cells within the total lymphocyte forward/side scatter gate, the CD4+/CD25HI cells (right panels) were gated and analyzed for Foxp3 expression (left panels).

Figure 3

DAB/IL2 transiently depletes all analyzed CD4+ T cell subsets. Whole blood was collected from 10 patients throughout the first cycle of DAB/IL2 and analyzed for CD4+/CD25HI/Foxp3+ co-expression by flow cytometry as described in the Figure 2 legend and Methods section. The absolute concentration of CD4+/CD25- (black), CD4+/CD25+ (red) and CD4+/CD25HI (green) T cells (A) and of CD4+/CD25HI/Foxp3- (black) and CD4+/CD25HI/Foxp3+ (red) T cells (B) were quantified by multiplying the percentage of anti-CD4, anti-CD25 and/or anti-Foxp3 fluorescence-positive cells within the lymphocyte forward/side scatter gate by the absolute lymphocyte concentration determined using an automated hematology analyzer. The percent control of each sample was calculated by dividing the absolute cell concentration on the indicated day of treatment with the cell concentration on day 0 prior to DAB/IL2 administration (× 100). Data are represented as averages ± standard error of the mean (n = 10 patients).

Figure 4

Reduction in the T cell depleting activity of DAB/IL2 during cycles 2–4 is associated with the development of anti-DAB/IL2 IgG. Whole blood was collected on the indicated days from patients P3 (A, B), P7 (C, D), P9 (E, F) and P16 (G, H) throughout four cycles of DAB/IL2 administration (each cycle indicated by an arrow). CD4+ (black), CD8+ (red), CD4+/CD25HI/Foxp3- (green) and CD4+/CD25HI/Foxp3+ (purple) T cells and monocytes (blue) were quantified as described in Figures 1 – 3 (A, C, E, H). Percent control of each sample was calculated by dividing the absolute cell concentration on the indicated day of treatment with the cell concentration on day 0 prior to DAB/IL2 administration (× 100). Plasma was isolated on the indicated days and analyzed for the presence of anti-DAB/IL2 IgG by ELISA. For the ELISA, data are presented as averages ± standard deviations (n = 5 per sample).

Figure 5

Flow cytometric scatter plots demonstrating the de novo appearance of MART1-, gp100- and tyrosinase-specific CD8+ T cells after one cycle of DAB/IL2. Whole blood was collected from patient P16 during cycle one of DAB/IL2 administration just prior to (day 0) and 21 days after the first dose of DAB/IL2. The peripheral blood mononuclear cells were isolated from the whole blood by Ficoll gradient centrifugation and stained with a PE-labeled anti-CD8 monoclonal antibody and the indicated APC-labeled tetrameric HLA-A2*0201/peptide conjugates.

Figure 6

De novo appearance of MART1-, gp100- and/or tyrosinase-specific CD8+ T cells in 4/7 HLA-A2*0201+ melanoma patients after one cycle of DAB/IL2. Whole blood was collected from patients P7 (A), P9 (B), P14 (C) and P16 (D) throughout four cycles of DAB/IL2 administration (each cycle indicated by an arrow). The peripheral blood mononuclear cells were isolated from the whole blood by Ficoll gradient centrifugation and stained with a PE-labeled anti-CD8 monoclonal antibody and APC-labeled tetrameric HLA-A2*0201/MART1 (black) or gp100 (red) or tyrosinase (green) peptide conjugates. The peripheral blood concentration of the indicated melanoma antigen-specific CD8+ T cells was quantified by multiplying the percentage of CD8+/tetramer+ cells within the lymphocyte forward/side scatter gate by the absolute lymphocyte concentration determined using an automated hematology analyzer. Data are represented as averages ± standard error of the mean (n = 10 patients). Patient P14 did not develop detectable tyrosinase- or gp100-specific CD8+ T cells but the green line (tyrosinase) is concealing the red line (gp100) (C).

Figure 7

Regression of hepatic, mesenteric and hilar melanoma metastases after DAB/IL2 administration. A. Patient P3 was scanned by combination PET/CT imaging 2 weeks prior to DAB/IL2 administration (pre-DAB/IL2) and after completing four 3-week cycles of DAB/IL2 (post-DAB/IL2). The brain, heart and bladder have normal accumulations of the PET tracer 18F-fluorodeoxyglucose but several areas of increased metabolism consistent with melanoma metastases resolved after DAB/IL2 administration. B. CT imaging of patient P5 revealed a large right hilar mass and a mesenteric mass that both decreased in size after DAB/IL2 administration.

Figure 8

Regression of subcutaneous, intramuscular and lymphatic metastases after DAB/IL2 administration. A. The right lower extremity of patient P8 was scanned by CT imaging 3 weeks prior to DAB/IL2 administration (pre-DAB/IL2) and after completing four 3-week cycles of DAB/IL2 (post-DAB/IL2). The white numbers in the lower left corner of each image indicate the distance (mm) above the superior aspect of the patella in order to provide matched images for comparison. B. CT imaging of patient P9 revealed a rapidly growing right inguinal mass that decreased in size 3 months after DAB/IL2 administration. A follow-up scan, 6 months after DAB/IL2 administration, revealed no further growth.

Figure 9

Stabilization of two right hilar masses in a 79-year old male after DAB/IL2 administration. PET imaging of patient P12 was conducted 1 month prior and 3 months after 2 cycles of DAB/IL2. Two discrete areas of hypermetabolism in the right hilum remained stable during this three month period.

Figure 10

Near complete response of widespread visceralmelanoma metastases after 4 cycles of DAB/IL2. A. Anterior/posterior views, PET. B. Lateral views, PET. C. CT imaging, liver. D. CT imaging, lungs. Combined PET/CT imaging of patient P14 revealed rapid progression of multiple melanoma metastases in the liver, both lungs, lymph nodes and the subcutaneous compartment (compare -6 months to -1 week). After 4 cycles of DAB/IL2, the liver metastases completely resolved and the lung metastases markedly regressed (compare -1 week to +3 months). Three months after completion of DAB/IL2, the residual lung metastases had completely resolved but a single enlarged peri-aortic lymph node persisted (red arrow). The increased 18F-fluorodeoxyglucose uptake in the brain, bladder and both kidneys are due to normal metabolism and are not reflective of metastases.

Figure 11

CD8+ T cell infiltration of residual HLA-A, B and C negative melanoma and evidence for vitiligo after DAB/IL2 administration. The residual peri-aortic mass in patient P14 was resected, formalin fixed and embedded in paraffin. A. Hematoxylin/eosin (H&E) staining of the mass revealed a mononuclear infiltrate that was confirmed to include CD8+ T cells by double immunohistochemistry using an anti-CD8 antibody (brown) and an anti-MART1 antibody (red). The counter stain used in the immunohistochemistry was hematoxylin and the control consisted of no primary antibody. B. HLA-A, B or C expression by cells in the pancreas (top; positive control), an unrelated melanoma metastasis (middle; positive control) and the residual peri-aortic melanoma metastasis resected from patient P14 as determined using a monoclonal antibody specific for a non-polymorphic portion of these HLA molecules. C. Photographs of patient P14's hair before and after DAB/IL2 administration revealed the complete loss of pigmentation.