Increased Susceptibility to Fungal Infection in Recipients with GVHD Is Not Due to Impaired Pathogen Clearance and Can Be Prevented by Co-Transfer of Donor CD4⁺CD25⁺ Regulatory T Cells

Abstract 822

Opportunistic fungal infections - in particular invasive aspergillosis (IA) – are still a cause of considerable morbidity and mortality in patients undergoing allogeneic bone marrow transplantation (BMT). Susceptibility is particularly high during neutropenia early after allo-BMT. However, this period is usually only temporary and nowadays largely manageable by prophylactic anti-fungal medication. Graft-versus-host disease (GVHD) represents another frequent, life-threatening complication after BMT. It is mediated by mature alloreactive donor T cells and leads to dysregulated pro-inflammatory cytokine secretion and target organ destruction. GVHD ≥ grade II has repeatedly been identified as independent risk factor for late onset IA. However, due to the simultaneous pharmacologic treatment of patients with GVHD, the contribution of GVHD itself to course and severity of a fungal infection is so far incompletely understood. To answer this question, we employed an MHC-mismatched murine allo-BMT model (C57BL/6 into BALB/c), in which lethally irradiated recipients received 2.5 × 10⁶ BM cells either alone (control group without GVHD) or together with 1 × 10⁶ splenocytes (development of GVHD within 28 d after BMT [clinical GHVD score on average 4.0 (range: 3–5.5)]). On d 28 after BMT half of the recipients within both groups were infected with Aspergillus fumigatus (A.f.) conidia (strain D141; 0.5 × 10⁶ i.v.) and all animals were subsequently monitored for survival. Whereas all non-infected and 60% of infected BM control animals survived the observation period of 30 d, all infected recipients with GVHD died within 10 d after infection. As expected, 90% of uninfected recipients with GVHD also died within 30 d, however, animals of this group survived significantly longer than their infected counterparts (p=0.017 for infected vs. uninfected GVHD group and p=0.004 for infected GVHD versus infected BM control group; n=5–9). Comparable results were obtained after i.t. infection of recipients with and without GVHD. To analyze pathogen clearance, BMT recipients with or without GVHD were infected, sacrificed 1 or 2 d later and CFU determined in lung (after i.t. or i.v. infection) as well as spleen, liver, kidney and brain (after i.v. infection). CFU decreased rapidly and comparably in all organs and in both recipient groups, indicating that pathogen clearance is not impaired in allo-BMT recipients with GVHD. However, leukocytes re-isolated from animals with GVHD reacted much stronger to in vitro stimulation with live germinated A.f. conidia than cells from BM controls, as indicated by a significantly increased release of TNF (10951 ± 4673 pg/ml vs 862 ± 256 pg/ml from splenocytes of GVHD and BM control animals, respectively; p= 0.045, n=4–5 animals/group; 7640 ± 3524 pg/ml vs. 2556 ± 620 pg/ml from liver leukocytes of GVHD and BM control animals, respectively; p=0.02, n=3 independent experiments). Similar results were obtained for IL-6 and IFN-γ. These data suggest that an unrestrained and dysregulated immune response to the fungus occurs in animals with GVHD. To further investigate this hypothesis, we transplanted an additional group with donor CD4⁺CD25⁺ Treg cells (0.25 × 10⁶/mouse), BM cells (2.5 × 10⁶) and splenocytes (1 × 10⁶), infected the mice with A.f. conidia and compared their survival rate to that of infected GVHD animals. Interestingly, whereas again all mice without Treg cell transfer (GVHD group) died after infection, 60% of the animals that received Treg cells survived the observation period of 30 d (p=0.011; n=10). Importantly, Treg cell transfer did not interfere with pathogen clearance. The in vitro immune response to live A.f. conidia of lymphocytes re-isolated from Treg recipients, however, as measured by TNF secretion, was significantly reduced and resembled that of BM controls (TNF from liver leukocytes in pg/ml: BM control: 1693 ± 374; GVHD: 8280 ± 3561; Treg: 1921 ± 642; p=0.02 BM vs GVHD and GVHD vs Treg; n=3 independent experiments).

In summary, our data indicate that an uncontrolled anti-fungal immune response rather than an impaired ability to deal with the pathogen underlies the increased mortality observed in murine allo-BMT recipients with GVHD. Furthermore, we show that co-transferred CD4⁺CD25⁺ donor Treg cells do not interfere with pathogen clearance but confine the anti-fungal immune response and thereby increase survival of infected animals.