Role of interferon-γ in immune-mediated graft failure after allogeneic hematopoietic stem cell transplantation

Abstract

Pathophysiology of graft failure (GF) occurring after allogeneic hematopoietic stem cell transplantation (HSCT) still remains elusive. We measured serum levels of several different cytokines/chemokines in 15 children experiencing GF, comparing their values with those of 15 controls who had sustained donor cell engraftment. Already at day +3 after transplantation, patients developing GF had serum levels of interferon (IFN)-γ and CXCL9 (a chemokine specifically induced by IFNγ) significantly higher than those of controls (8859±7502 vs. 0 pg/mL, P=0.03, and 1514.0±773 vs. 233.6±50.1 pg/mlL, P=0.0006, respectively). The role played by IFNγ in HSCT-related GF was further supported by the observation that a rat anti-mouse IFNγ-neutralizing monoclonal antibody promotes donor cell engraftment in Ifngr1^-/-mice receiving an allograft. In comparison to controls, analysis of bone marrow-infiltrating T lymphocytes in patients experiencing GF documented a predominance of effector memory CD8⁺ cells, which showed markers of activation (overexpression of CD95 and downregulation of CD127) and exhaustion (CD57, CD279, CD223 and CD366). Finally, we obtained successful donor engraftment in 2 out of 3 children with primary hemophagocytic lymphohistiocytosis who, after experiencing GF, were re-transplanted from the same HLA-haploidentical donor under the compassionate use coverage of emapalumab, an anti-IFNγ monoclonal antibody recently approved by the US Food and Drug Administration for treatment of patients with primary hemophagocytic lymphohistiocytosis. Altogether, these results suggest that the IFNγ pathway plays a major role in GF occurring after HSCT. Increased serum levels of IFNγ and CXCL9 represent potential biomarkers useful for early diagnosis of GF and provide the rationale for exploring the therapeutic/preventive role of targeted neutralization of IFNγ.

Figure 1.

Cytokine/chemokine profile. Serum levels of interferon (IFN)-γ (A), CXCL9 (B), CXCL10 (C), and sIL2Rα (D) in patients who either did (red line) or did not (blue line) experience graft failure (GF). All graphs represent Mean and Standard Error of Mean for each variable. HSCT: hematopoietic stem cell transplantation.

Figure 2.

Cytokine/chemokine profile. Serum levels of TNFα (A), CXCL10 (B), sCD163 (C), IL6 (panel D). Red line: patients who experience graft failure (GF); blue line: controls. All graphs represent Mean and Standard Error of Mean for each variable. HSCT: hematopoietic stem cell transplantation.

Figure 3.

Immunohistochemistry evaluation of bone marrow (BM) specimens in a patient experiencing graft failure (Pt #4). (A) Hematoxylin & eosin (H&E) staining of a BM specimen at 4X magnification. (B) Evaluation of erythroid colony spreading by glycophorin staining (10X). (C) Megakaryocyte distribution evaluated by CD61 expression (10X). (D) H&E staining at 40X showing apoptotic events. (E) H&E staining revealing stromal damage and edema development (40X). Characterization of the macrophage population by CD68 (F) and CD163 (G) staining (40X). Characterization and distribution of T lymphocytes by analysis of CD3 (H), CD4 (I), and CD8 (J) expression (10X).

Figure 4.

Immunohistochemistry characterization of bone marrow (BM) in patients who either did or did not experience graft failure (GF). (A) Comparison of absolute number of CD3⁺, CD4⁺, CD8⁺, CD68⁺, TIA-1⁺, perforin⁺ and granzyme⁺ cells in BM of GF patients and controls (CTRL). The total number of positive cell for each marker was counted in five fields per sample under 20-fold magnification and reported as Mean±Standard Deviation. (B) Percentages of CD68⁺ cells with hemophagocytic activity (i.e. showing cellular fragments, erythrocytes and lipid vacuoles in their cytoplasm) in BM of GF patients and CTRL. ^*P<0.05; ^**P<0.01; ^***P<0.001.

Figure 5.

Immuno-characterization of the T lymphocytes present in bone marrow aspirates of patients who either did or did not experience graft failure (GF). (A) Flow cytometry analysis of CD4⁺ and CD8⁺ population in patients with GF and controls (CTRL). Distribution of naïve (CD45RA⁺/CCR7⁺), central memory (CD45RO⁺/CCR7⁺), effector memory (CD45RO⁺/CCR7⁻), effector terminal (CD45RA⁺/CCR7⁻), and NK-T (CD3⁺/CD56⁺) subsets in CD4⁺ (B) or CD8⁺ (C) T cells. Activation and exhaustion profile in both the CD4⁺ and CD8⁺ population by the analysis of CD95 (D), CD127 (E), and CD57 (F). (A, D, E, and F) Each patient or CTRL is represented by a symbol and a horizontal line marks the median. (B and C) The average (+) and Median±Standard Deviation are shown. ^*P<0.05; ^**P<0.01; ^***P<0.001; ^****P<0.0001.

Figure 6.

Successful hematopoietic stem cell transplantation (HSCT) chimerism in interferon (IFN)-γR1^−/− mice correlates with low IFNγ activity; circulating CXCL9 levels is a biomarker of in vivo IFNγ activity. Ifngr1^−/− mice (expressing the Ly5.2 congenic marker) were intravenously (i.v.) infected with 1,106 CFU of Bacillus Calmette–Guérin (BCG) (strain Pasteur 1173P2). After 14, 20, 28, 35 and 42 days mice were treated i.v. with 100 mg/kg of an isotype control (n=5) or the anti-mIFNγ, XMG1.2 (n=5). At day 21, mice were infused with bone marrow from Ifngr^+/+ mice, expressing the Ly5.1 marker, after mild irradiation (550 rads). Chimerism, assessed by determining the surface expression of Ly5.1 and Ly5.2 on lymphocytes, was analyzed by flow cytometry at different time points after HSCT treatment. IFNγ levels were quantified at different time points by ELISA using the Luminex technology. (A) Graph represents the super-imposition of the chimerism (black straight line) and the IFNγ levels (gray dotted line) in the isotype control treated mice. (B) Graph represents the chimerism determined in mice treated with the isotype control (black straight line) or with the XMG1.2 (gray straight line) mAbs. (C) Ifngr1^−/− mice were i.v. infected with 1.106 CFU of BCG (strain Pasteur 1173P2). After 14, 20, 28, 35 and 42 days mice were treated i.v. with 100 mg/kg of an isotype control (black straight line; n=5) or the anti-mIFNγ, XMG1.2 (gray straight line; n=5). At day 21, mice were transplanted with bone marrow from Ifngr^+/+ mice, expressing the Ly5.1 marker, after mild irradiation (550 rads). At different time points post-BCG infection, circulating CXCL9 levels were quantified by ELISA using the Luminex technology. Ab: antibody.