Abstract

Abstract 43

We have previously shown that TNF Receptor 1 (TNFR1) levels early after myeloablative hematopoietic cell transplant (HCT), especially following unrelated donor (URD) HCT, predict later graft-versus-host disease (GVHD), non-relapse mortality (NRM), and overall survival (OS). Given this data we hypothesized that augmenting standard GVHD prophylaxis with the TNF-inhibitor etanercept would protect against GVHD and NRM while improving survival. We therefore conducted an IRB-approved Phase II trial at two centers to determine whether prophylactic TNF blockade with etanercept decreases the rates of acute GVHD and day 100 NRM seen after high-risk allogeneic matched (n=68, 71%) or single antigen mismatched URD (n=25, 26%) or single antigen mismatched related donor (RD) (n=3, 3%) HCT. Follow-up data are available for 96/100 patients (pts) enrolled between 2005 and 2009. HLA matching used high-resolution DNA-based techniques for HLA-A, B, C, and DR. Patients received tacrolimus, methotrexate (5 mg/m2 on days 1, 3, 6, 11) and etanercept (0.4 mg/kg, max 25 mg) twice weekly from the start of conditioning to day 56. All pts received myeloablative regimens; either TBI/cyclophosphamide (TBI, n=29), fludarabine and 12.8 mg/kg IV busulfan (FluBu4, n=31) or other myeloablative chemotherapy regimens, either BCNU/etoposide/cyclophosphamide or busulfan/cyclophosphamide (BCNU/Busulfan, n=36). The median age of the pts was 45y (range 2-61y).

To determine whether the administration of etanercept effectively reduced post-transplant TNFR1 levels we compared day 7 post-transplant TNFR1 levels in the 96 study pts to 132 control pts who received standard GVHD prophylaxis of tacrolimus/methotrexate and were matched for age, conditioning regimen, degree of HLA-match, and disease status. The median day 7 TNFR1 levels in study pts was 2518 pg/ml, significantly lower than the median day 7 level of 3529 pg/ml in control pts (p<0.001), thus confirming that etanercept administration is associated with lower TNFR1 levels early post-transplant. Similar statistically significant results were seen with mean TNFR1 levels. We then tested whether day 7 TNFR1 levels were associated with outcomes in etanercept treated pts. For each doubling of TNFR1 levels, there was an increased risk for NRM (HR 2.0, p<0.001) and death (HR 1.5, p<0.06); the association with NRM remained significant after adjusting for age, gender, risk group, and donor (p=0.002). Interestingly, as shown in the Table, we observed that the effect of etanercept on day 7 TNFR1 levels differed statistically according to the myeloablative conditioning regimen administered (p<0.001). Patients who received FluBu4 had the lowest median day 7 TNFR1 levels and also experienced the best outcomes with low rates of grade 2-4 GVHD after high-risk HCT (39%, p=0.04), no deaths in the first 100 days (p<0.001), and outstanding 2 year survival rates (70%, p=0.03), compared to other study pts.

PATIENT OUTCOMES BY CONDITIONING REGIMEN

 Sample Size Median Day 7 TNFR1 Levels (pg/mL) Mean Day 7 TNFR1 Levels (pg/mL) Day 100 Grade 2-4 GVHD Day 100 Grade 3-4 GVHD Day 100 NRM 2 yr NRM 2 yr OS 
OVERALL 96 2518 3053 53% 23% 16% 34% 55% 
FluBu4 31 1735 1897 39% 11% 0% 12% 70% 
BCNU/Busulfan 36 2552 3009 61% 25% 14% 31% 58% 
TBI 29 3907 4226 62% 32% 34% 55% 38% 
 Sample Size Median Day 7 TNFR1 Levels (pg/mL) Mean Day 7 TNFR1 Levels (pg/mL) Day 100 Grade 2-4 GVHD Day 100 Grade 3-4 GVHD Day 100 NRM 2 yr NRM 2 yr OS 
OVERALL 96 2518 3053 53% 23% 16% 34% 55% 
FluBu4 31 1735 1897 39% 11% 0% 12% 70% 
BCNU/Busulfan 36 2552 3009 61% 25% 14% 31% 58% 
TBI 29 3907 4226 62% 32% 34% 55% 38% 

We have evidence that etanercept effectively lowers TNFR1 levels at day 7 post-transplant, a time point associated with GVHD and HCT outcomes. However, the degree of TNFR1 level reduction appears to depend in part on the conditioning regimen. The median day 7 TNFR1 levels were statistically different across conditioning regimens, and those with low TNFR1 levels have improved outcomes (FluBu4>BCNU/Busulfan>TBI). In fact, for the FluBu4 pts in particular, preliminary data suggest superior outcomes, including low rates of GVHD and remarkably high 2 year OS for high-risk transplant pts, without use of thymoglobulin. After highly intense conditioning regimens, such as TBI-based, TNFR1 levels remained high despite TNF reduction with etanercept and outcomes are similar to those expected with standard GVHD prophylaxis. With the less toxic, but still myeloablative FluBu4 conditioning regimen, however, the addition of etanercept appeared to lower TNFR1 levels to levels that were protective against NRM. Thus, the effectiveness of etanercept may depend on the conditioning regimen, warranting a randomized trial to explore this interaction further.

Disclosures:

Off Label Use: Etanercept - for GVHD prophylaxis.

Author notes

*

Asterisk with author names denotes non-ASH members.