Impact of Early Blood Transfusion After Kidney Transplantation on the Incidence of Donor-Specific Anti-HLA Antibodies

I. Ferrandiz; N. Congy-Jolivet; A. Del Bello; B. Debiol; K. Trébern-Launay; L. Esposito; D. Milongo; G. Dörr; L. Rostaing; N. Kamar

Disclosures

American Journal of Transplantation. 2016;16(9):2661-2669. 

In This Article

Results

Blood Transfusion

A total of 250 patients (64.1%) received a blood transfusion within the first year after transplant. All received an RBC transfusion, 22 patients received a transfusion of platelets and 11 patients received a fresh plasma infusion. A total of 224 patients received only RBCs; 15 patients received RBCs plus platelet transfusions; 4 patients received RBCs plus fresh plasma infusions; and 7 received RBCs, platelets and fresh plasma infusions.

The median time between kidney transplantation and the first blood transfusion was 3 days (range: 0–344 days). Within the first month after transplantation, 94.8% of patients required a blood transfusion. The median number of transfusions was one per patient (range of one to seven per patient): 63.2% of patients received only one transfusion, and 94.4% received fewer than three transfusions. The mean hemoglobin level at blood transfusion was 8.4 ± 0.7 g/dL. The causes for blood transfusion were a hemoglobin level <10 g/dL within the first postoperative days (n = 220, 88%), a hemoglobin level <10 g/dL in patients with impaired kidney function (n = 10, 4%), a severe infection (n = 10, 4%) or an acute hemorrhage (n = 10, 4%).

Comparison of the Patient Characteristics for Those who did and did not Receive a Blood Transfusion

Comparisons between patients who did and did not receive a blood transfusion are presented in Table 1 . Male patients needed significantly fewer transfusions. At transplantation, hemoglobin level was significantly lower in patients who required a blood transfusion (11.8 ± 1.2 g/dL vs. 12.6 ± 1.2 g/dL, p < 0.0001). More patients who had a blood transfusion also received an induction therapy at transplantation (75.6% vs. 60.7% with no transfusion, p = 0.003). The overall proportion of patients who received a T cell–depleting induction therapy was low, but the proportion was significantly higher among patients who needed a transfusion (12.4% vs. 3.6%, respectively, p = 0.003). Patients who were given induction therapy were more frequently women, had less often received a kidney from a deceased donor and had a lower hemoglobin level at transplantation (Table S1).

With respect to the initial immunosuppressive therapy, the proportion of patients given tacrolimus was significantly lower in those that needed a transfusion. Conversely, the proportion of patients given cyclosporine A or a de novo mTOR inhibitor was significantly higher in the transfusion group.

Incidence of anti-HLA Antibodies and Outcomes

During the first year post-transplantation, 34 patients (8.7%) developed de novo anti-HLA antibodies, 32 (12.8%) in the transfusion group and two (1.4%) from the nontransfusion group (p < 0.0001) (Table 2). Nineteen patients (4.9%) developed de novo DSAs, 18 (7.2%) in the transfusion group and only one (0.7%) in the nontransfusion group (p < 0.0001). Of these 18 patients from the transfusion group, eight developed both anti–class I and anti–class II antibodies, and the other 10 patients developed only anti–class II antibodies. The median MFI of the immunodominant antibodies was 6000 (range: 1000–19 000). The median number of DSAs per patient was one (range of one to three per patient). Only one patient from the nontransfusion group developed a de novo class II anti-HLA antibody; the MFI was 12 000.

Within the first year after transplantation, the incidence of biopsy-proven AMR was significantly higher in the transfusion group (6%) compared with the nontransfusion group (1.4%; p = 0.04). In the transfusion group, 15 of the 18 patients who developed DSAs experienced AMR, as did the patient who developed DSAs in the nontransfusion group. All patients were treated by plasma exchange, rituximab and intravenous immunoglobulin. During the first posttransplantation year, only one graft loss related to severe cellular acute rejection occurred in a transfused patient; this was at 249 days after transplantation. No anti-HLA antibodies were detected in this patient at graft loss. Five patients (1.3%) died in the transfusion group within the first year after transplantation; the causes of death were hemorrhagic shock (n= 1), septic shock (n= 3) and acute respiratory distress (n= 1). No patient died in the nontransfusion group.

Immunosuppression in Patients who Received a Blood Transfusion

At transplantation, patients who developed a DSA at 1 year had received significantly more frequent anti-CD25 mAb induction therapies (89%) compared with those who did not develop a DSA (61.2%; p= 0.02). Conversely, in transfused patients, the use of polyclonal antibody induction therapy did not significantly affect the occurrence of DSAs at 1 year. In addition, patients who developed DSAs at 1 year were significantly less frequently treated by tacrolimus at first transfusion (38.9%) compared with those who did not develop DSAs (78%; p= 0.0007). Conversely, patients who developed DSAs at 1 year were significantly more frequently treated by cyclosporine A at first transfusion (55.5%) compared with those who did not develop DSAs (19.4%; p= 0.00013). The following, however, did not differ between patients who developed or did not develop DSAs at 1 year: tacrolimus trough levels, the proportions of patients who had a tacrolimus trough level of <5 ng/mL during the first year, cyclosporine A concentration at 2 h after intake (C2), the proportions of patients that had a C2 level <500 ng/mL during the first year, the proportions of patients with trough levels of mTOR inhibitor <5 ng/mL during the first year and the proportions of patients receiving MPA and steroids (and their respective daily doses) ( Table 3 ). Within the first year after transplant, among the 18 transfused patients who developed DSAs, one patient had a BK virus replication without BK virus nephropathy, two were suspected for nonadherence, four had been hospitalized for bacterial infection, none had a cytomegalovirus replication and 10 patients had at least one low exposure to immunosuppressants (Table S2).

Predictive Factors for the Occurrence of de novo DSAs Within the First Year After Transplantation

Predictive factors for development of de novo DSAs within the first year after transplantation were determined. Results from univariate analyses are presented in Table 4. The following variables were included in the multivariate analyses: number of HLA A/B/DR/DQ mismatches, blood transfusion (vs. no blood transfusion), the total number of packed RBC units given, the use of anti–IL-2 receptor blockers as an induction therapy, the use of tacrolimus (vs. cyclosporin A) and a cellular rejection episode (vs. no acute rejection) (Table 5). A blood transfusion (odds ratio [OR] 8.94; 95% confidence interval [CI] 1.13–70.76; p = 0.04), the occurrence of a cellular acute rejection episode (OR 3.27; 95% CI 1.06–10.05; p = 0.04) and the number of HLA A/B/DR/DQ mismatches (OR 1.53; 95% CI 1.01–2.33; p = 0.04) were identified as predictive factors for the formation of de novo DSAs. Conversely, the use of tacrolimus (vs. cyclosporine A) had a significantly lower likelihood of being associated with de novo DSAs (OR 0.17; 95% CI 0.06–0.47; p = 0.0006).

Predictive Factors for the Occurrence of AMR Within the First Year After Transplantation

Predictive factors for the occurrence of AMR within the first year after transplantation were determined. Results from the univariate analyses are presented in Table 6. The following variables were included in the multivariate analyses: number of HLA A/B/DR/DQ mismatches, blood transfusion (vs. no blood transfusion), DSAs at 1 year (vs. no DSAs), anti-HLA antibodies at 1 year (vs. no anti-HLA antibodies), an induction therapy (vs. no induction therapy), the use of anti–IL-2 receptor blockers as an induction therapy, the use of tacrolimus (vs. cyclosporin A) and a cellular rejection episode (vs. no acute rejection). The presence of DSAs (OR 30.43; 95% CI 9.68–95.65; p < 0.0001) was the sole predictive factor for the development of AMR. Blood transfusion was not a predictive factor for AMR (OR 1.98; 95% CI 0.4–9.85; p = 0.4).

processing....