Background

Advances in immunosuppressive drugs have improved the incidence of acute rejection and the prognosis of transplanted kidneys, but antibody-mediated rejection (ABMR) caused by de novo donor-specific antibody (DSA) to human leukocyte antigen (HLA) remains difficult to treat and is a major cause of transplant renal dysfunction [1,2,3,4,5,6]. However, the association between the development of de novo DSA and ABMR is unclear, as not all DSA-positive patients develop ABMR. Additionally, no standard treatment for ABMR due to de novo DSA has been established.

In 2018, annual anti-HLA antibody screening tests were approved for renal transplant recipients in Japan. Herein, we report the prevalence of de novo DSA and evaluate the association of de novo DSA with functional graft outcomes in living-donor renal transplant recipients at our hospital.

Methods

Study design and participants

The present study was performed in accordance with the guidelines of the Declaration of Helsinki and was approved by the research ethics committee of Kyoto University (#R2485-1).

Of the 110 patients who underwent living-donor renal transplantation from 1980 to 2019, 80 patients who were screened for anti-HLA antibody were retrospectively reviewed. One patient who had preformed DSA was excluded. The other patient was excluded because he had a myocardial infarction soon after surgery and died early. Patient characteristics such as age, sex, ABO-compatibility, cause of renal dysfunction, and history of liver transplantation were investigated based on the clinical records. The patient characteristics are shown in Table 1. The mean age at transplantation was 43.2 ± 14.6 years. There were 43 men (53.8%) and 68 ABO-compatible cases (85.0%). IgA nephropathy was the most common cause of renal dysfunction (n = 20, 25.0%). There were one secondary renal transplant cases, and five patients had a history of liver transplantation. The median follow-up period was 101 months (range 17–487 months).

Table 1 Characteristics of renal transplantation recipients
Full size table

The estimated glomerular filtration rate and the monthly trough concentration of cyclosporin A and tacrolimus were compared using the Mann–Whitney U test. The functional graft survival was estimated by the Kaplan–Meier method and compared by the log-rank test. Statistical significance was set at P < 0.05. All analyses were completed using JMP® software (JMP pro version 15; SAS Institute Inc., Cary, NC, USA).

Results

Anti-HLA antibody was detected in 14 patients (17.5%), comprising DSA in eight patients (10.0%) (Fig. 1). Functional graft loss occurred in two (25%) of the eight patients who were positive for DSA, none of the six patients who were positive for non-DSA anti-HLA antibody, and none of the 66 patients who were negative for anti-HLA antibody (P = 0.0419, log-rank test, Fig. 2). The mean estimated glomerular filtration rate at the time of the anti-HLA antibody test was 45.1 ± 14.4 mL/min/1.73m2 in the 66 patients with no anti-HLA antibody, while it was 35.0 ± 11.5 mL/min/1.73m2 in the eight patients with de novo DSA (P = 0.0702) and 39.3 ± 15.3 mL/min/1.73m2 in the six patients with non-DSA anti-HLA antibody (P = 0.3921). The mean monthly cyclosporin A trough concentration for the past year from the anti-HLA antibody test was 59.2 ± 24.8 ng/ml in the seven patients with no anti-HLA antibody, while it was 61.9 ± 12.5 ng/ml in the five patients with de novo DSA (P = 0.5670) and 36.3 ± 9.0 ng/ml in a patient with non-DSA anti-HLA antibody (P = 0.3921) (Fig. 3). The mean monthly tacrolimus trough concentration for the past year from the anti-HLA antibody test was 4.62 ± 1.20 ng/ml in the 55 patients with no anti-HLA antibody, while it was 4.09 ± 1.10 ng/ml in the three patients with de novo DSA (P = 0.0027) and 4.21 ± 1.14 ng/ml in the four patients with non-DSA anti-HLA antibody (P = 0.0722) (Fig. 4).

Fig. 1
figure 1

Anti-HLA antibody screening results. HLA: human leukocyte antigen; DSA: donor-specific antigen

Full size image
Fig. 2
figure 2

Kaplan–Meier curve of renal transplant survival rate. DSA: donor-specific antigen

Full size image
Fig. 3
figure 3

A Cyclosporin A troughs over the immediate past year prior to DSA detection. B Comparison of cyclosporin A troughs over the immediate past year prior to DSA detection. CyA: cyclosporin A; DSA: donor-specific antigen

Full size image
Fig. 4
figure 4

A Tacrolimus troughs over the immediate past year prior to DSA detection. B Comparison of tacrolimus troughs over the immediate past year prior to DSA detection. DSA: donor-specific antigen; mTOR: mammalian target of rapamycin; CNI: calcineurin inhibitor

Full size image

Table 2 summarizes the characteristics of the eight patients who were positive for de novo DSA. All eight patients received an ABO-compatible renal transplant. Delayed graft function occurred in two patients. De novo DSA was detected at a median of 141.5 months (range 18–275 months) after transplantation. In all cases, the de novo DSA was HLA class II (DQ only in three patients, DR only in three, and DQ + DR in two). Six patients (75%) had a mean fluorescence intensity (MFI) of 10,000 or more. The transplanted renal function was lost in two patients (25%). Four patients (50%) did not show any signs of rejection, while the remaining four were screened for anti-HLA antibody due to elevated serum creatinine levels. Renal biopsy was performed in all eight patients, and five (62.5%) were pathologically diagnosed with ABMR (subclinical ABMR in two patients, and clinical ABMR in three). Table 3 shows the pathological findings (Banff classification).　Immunosuppression was enhanced using steroid pulse therapy, plasmapheresis, rituximab, intravenous immunoglobulin, and dose increase in maintenance immunosuppressants.

Table 2 Clinical characteristics of DSA-positive recipients
Full size table
Table 3 Banff classification of DSA-positive recipients
Full size table

Two patients with loss of transplanted renal function are described below.

Case 3 was a 27-year-old man (body weight 80 kg) with chronic glomerulonephritis who received a living-donor renal transplant from her mother (Fig. 5). The maintenance immunosuppressant therapy was a combination of cyclosporin A (1.5 mg/kg), prednisolone (5 mg/body), and azathioprine (0.9 mg/kg). Twenty years after the renal transplantation, his serum creatinine was stable at 1.5–1.6 mg/dl. A graft kidney biopsy was performed because DSA was detected on blood sampling. Pathological findings indicated ABMR. The azathioprine was changed to mycophenolate mofetil (MMF) (2000 mg/body), and his MFI was halved 1 year later, but his serum creatinine increased after that. His serum creatinine level decreased temporarily with steroid pulse therapy (methylprednisolone 500 mg/body for 3 days). At that time, the patient did not want additional treatment. However, his serum creatinine increased again 4 months later, and hemodialysis was reintroduced at 23 years after the kidney transplantation.

Fig. 5
figure 5

Clinical course and change in serum creatinine level in case #3. Case 3 is a 27-year-old man with chronic glomerulonephritis who received an ABO-compatible transplant with her mother as the donor. CyA: cyclosporin A; PSL: prednisolone; AZA: azathioprine; MMF: mycophenolate mofetil; cABMR: chronic antibody-mediated rejection; DSA: donor-specific antigen; HD: hemodialysis

Full size image

Case 4 was a 49-year-old woman (body weight 40 kg) with an unknown primary disease who received a living-donor renal transplantation from her husband (Fig. 6). Maintenance immunotherapy was a combination of cyclosporin A (2 mg/kg), prednisolone (5 mg/body), and MMF (1000 mg/body). Her serum creatinine was stable at about 1.0 mg/dl for 11 years after renal transplantation, but then gradually increased. The MMF was changed to mizoribine (2.5 mg/kg) when she developed herpes zoster. After that, she had an elevated serum creatinine level and proteinuria of 3 g/day. DSA was detected and renal biopsy revealed ABMR. The patient did not respond to any immunosuppression enhancement including steroid pulse therapy (methylprednisolone 250 mg/body on the first day, 100 mg/body on the second day, 100 mg/body on the third day), two sessions of plasmapheresis, rituximab administration (250 mg/body), intravenous immunoglobulin (2 g/kg), and everolimus (1.5 g/body) administration. Hemodialysis was restarted at 13 years after the renal transplant.

Fig. 6
figure 6

Clinical course and change in serum creatinine level in case #4. Case 4 is a 49-year-old woman with an unknown primary disease who received an ABO-compatible transplant with her husband as the donor. CyA: cyclosporin A; PSL: prednisolone; MMF: mycophenolate mofetil; cABMR: chronic antibody-mediated rejection; DSA: donor-specific antigen; PE: plasma exchange; IVIG: intravenous immunoglobulin; HD: hemodialysis

Full size image

Discussion

The pathophysiology of ABMR is that the produced DSA attacks the vascular endothelium of the transplanted kidney, resulting in tissue damage [7]. ABMR is still one of the main causes of renal transplant loss, but it is often difficult to treat at the time of diagnosis because it develops subclinically and tissue damage progresses.

DSA is broadly divided into preformed DSA that exists in serum before transplantation and de novo DSA that is newly produced after transplantation [8, 9]. Preformed DSA is caused by pregnancy, past blood transfusions, and organ transplants. Patients with preformed DSA are at high risk of acute ABMR; even if desensitization therapy is performed before transplantation, the incidence of ABMR is 1–10% [10]. Patients with de novo DSA develop ABMR later than patients with preformed DSA [8, 9] and have a reduced long-term survival rate [11].

The reported incidence of de novo DSA varies, but is about 2–10% at 1 year after renal transplantation, increases by 2% per year and reaches about 10–40% at 4–5 years after transplantation [11,12,13,14,15]. The average time of de novo DSA appearance is reportedly 3.8–68 months after transplantation [12, 15, 16]. In the present study, de novo DSA was detected at a median time of 141.5 months after transplantation (range 18–275 months), which is longer than previously reported, but we consider that DSA was probably present before it was detected. In addition, the de novo DSA is reportedly predominantly class II antibody [11, 12, 16], which is a risk factor for ABMR [11, 17]. Similarly in the present study, the de novo DSA comprised class II antibodies in all eight patients.

It has been reported that the occurrence of de novo DSA is more common when the blood concentration of tacrolimus fluctuates greatly [18] and when the trough value of tacrolimus is 8 ng/ml or less [19]; thus, such patients should be monitored for the appearance of DSA. This may lead to the early detection of patients at risk of rejection due to inadequate immunosuppression. In addition, monitoring for DSA may improve the prognosis of the transplanted kidney through early diagnosis and treatment at the subclinical ABMR stage before the clinical appearance of renal damage [5, 20, 21]. In the present study, subclinical ABMR was diagnosed in case 3, but as 20 years had passed since the renal transplantation and the MFI had a markedly high value of 21,566, it is considered that DSA was present earlier. In addition, although cyclosporine A was used for CNI in many cases in the past, tacrolimus was used in many cases in recent years, and de novo DSA was frequently observed in patients using cyclosporine A. The target trough for cyclosporine A was less than 100 ng/ml, and the target trough for tacrolimus was 4–6 ng/ml, but the tacrolimus trough concentration was significantly lower in patients with de novo DSA than in patients with no anti-HLA antibody. (4.09 ± 1.10 ng/ml vs. 4.62 ± 1.20 P = 0.0027) (Fig. 4). Therefore, more stringent control of blood levels may be necessary to reduce the occurrence of de novo DSA.

Although there is no established treatment for ABMR, the treatment options include steroid pulse therapy, rabbit antithymocyte globulin administration, antibody removal using plasma exchange, rituximab administration, immunoglobulin therapy, bortezomib administration, and eculizumab administration [22]. Steroid pulse therapy alone has a low therapeutic effect for ABMR and should be used in combination with other drugs [23]. Plasma exchange and combined therapy comprising plasma exchange and immunoglobulin administration are reportedly effective treatments for ABMR [22, 24]. ABMR with de novo DNA is considered more resistant to treatment than ABMR with preformed DSA [25]. Our treatment strategy for ABMR is steroid pulses followed by plasmapheresis, rituximab, intravenous immunoglobulin, and intensified maintenance immunotherapy. In the present study, two of the eight patients with ABMR did not respond to treatment and eventually lost the transplanted renal function.

The present study had several limitations. This was a retrospective study with a small sample size, the timing of renal transplantation varied, and patients who had already lost transplanted renal function were not included. In addition, the follow-up period after anti-HLA antibody screening was short. It is necessary to accumulate more cases and continue with long-term follow-up to determine whether early intervention for ABMR improves the prognosis of the transplanted kidney.

Conclusions

In the present study, the incidence of anti-HLA antibody development was 17.5%, including the development of DSA in 10.0%. Two (25%) of the eight patients with de novo DSA eventually had graft loss. Most immunosuppressive enhancement therapies were not effective in decreasing the DSA levels. As there is no effective treatment established for the reduction of DSA, careful immunosuppressive management is essential to prevent the development of de novo DSA.