Elevations of Tumor Necrosis Factor-Receptor-1 at Day 7 and Acute Graft-Versus-Host Disease After Allogeneic Hematopietic Cell Transplantation with Nonmyeloablative Conditioning.

Acute graft-versus-host disease (GVHD) has remained a significant cause of nonrelapse mortality after allogeneic hematopoietic cell transplantation (HCT) with nonmyeloablative conditioning. The role of tumor necrosis factor-alpha (TNF-α) in the biology of acute GVHD following nonmyeloablative conditioning has not been studied thus far. Here, we measured TNF receptor 1 (TNFR1) as a surrogate marker for TNF-α in 106 patients before the start of the conditioning regimen (baseline) and 7 days after allogeneic HCT following nonmyeloablative conditioning.

The nonmyeloablative regimen consisted of 2 Gy total body irradiation (TBI) alone (n=15), 2 Gy TBI plus fludarabine 90 mg/m² (n=18). Postgrafting immunosuppression combined mycophenolate mofetil (MMF) with a calcineurin inhibitor for all patients. Blood samples were prospectively collected before the start of the conditioning regimen, then on days 7, 14, 21, 28, 35, 42, 49, 56, 63, 70, 77, 84, 91, and 98 after HCT, and then generally once every 2 weeks up to day 180. The serum component of each blood sample was separated and frozen for later analysis on the day of sample acquisition. TNFR1 serum concentration was retrospectively assessed using a cytokine enzyme-linked immunoabsorbent assay (R&D, Minneapolis, MN) according to the manufacturer's protocol.

TNFR1 levels increased significantly from baseline to day 7 after nonmyeloablative HCT (P<0.0001). Patients conditioned with 4 Gy TBI had higher TNFR1 day 7/baseline ratio than those conditioned with 2 Gy TBI (median 1.65 versus 1.25; P=0.01). Median time for diagnosis of grade II-IV acute GVHD was 38 (range, 4-341) days, with only 1 patient experiencing acute GVHD before day 7. In a multivariate Cox model, high TNFR1 day 7/baseline ratio (modeled as a continuous linear variable) was the only factor statistically significantly associated with a higher risk of grade II-IV (HR 2.2, P=0.01) and grade III-IV (HR 2.9, P=0.007) acute GVHD. There was also a suggestion for an association between a high incidence of grade II-IV acute GVHD and HLA-disparity between donor and recipient, older patient age, and female donor to male recipient, although these factors did not reach statistically significance perhaps because of the relatively small number of patients analyzed. Interestingly, day 7 TNFR1 levels were not statistically significantly associated grade II-IV acute GVHD (P=0.07), suggesting that TNFR1 day 7/baseline ratio predicts better for acute GVHD than TNFR1 day 7 alone. To further analyze the role of TNF in acute GVHD after nonmyeloablative conditioning, we compared TNFR1 levels at onset of acute GVHD (median, day 38) in a subgroup of 20 patients with grade II-IV acute GVHD, versus TNFR1 levels around day 38 after HCT in a subgroup of 20 patients who never experienced grade II-IV acute GVHD. TNFR1 levels were significantly higher in patients with grade II-IV acute GVHD than in those without (median 7,119 versus 3,140 pg/mL, P=0.001). One- and three-year overall survival rates were 65% and 45%, respectively. In multivariate analysis, unrelated donor (P=0.01), high disease risk (P=0.05), and higher patient age (P=0.03) were each associated with a higher risk of mortality, while TNFR1 day 7/baseline ratio (modeled as a continuous linear variable) was not (P=0.75).

Our data suggest that nonmyeloablative conditioning induces the generation of TNF-α, and that the magnitude of TNF-α generation depends on the conditioning intensity (2 Gy versus 4 Gy TBI). Further, assessment of TNFR1 levels before and on day 7 after nonmyeloablative HCT provided useful information on subsequent risk of experiencing acute GVHD.

SHB and EW contributed equally to the work.

No relevant conflicts of interest to declare.