ABO-incompatible renal transplantation in developing world – crossing the immunological (and mental) barrier

Introduction

Kidney transplantation is the best form of renal replacement therapy (RRT) for end-stage renal disease (ESRD) patients.[1] Only 10% of ESRD patients receive any form of RRT in India and only 2% undergo renal transplantation.[2] As per the Indian Chronic Kidney Disease (CKD) Registry, 39% of CKD 5 patients were on RRT and 2% were being worked up for renal transplant.[3] Renal transplant options for ESRD patients are limited. Deceased donor transplantation is still in its nascent stage. ABO incompatibility is an important limiting factor amongst willing donors, leading to rejection of approximately 35% donors.[4] In such a scenario, options are limited. One can opt for paired kidney exchange transplantation. Although it is being done at individual centers, there are no regional bodies to facilitate such exchanges between different centers and there is no national registry. This limits the number of such exchanges significantly. Also, many recipients are reluctant to accept kidney from donor outside their family in exchange of their own. ABO-incompatible (ABOi) transplant comes across as a viable alternative in such a scenario.

Although worldwide popularity of ABOi renal transplant is increasing, experience from developing world is limited. Major concerns are high cost, risk of antibody-mediated rejection and increased incidence of posttransplant infections. We hereby present experience of ABOi renal transplant at our center, a tertiary care center in north India.

Results

Table 1 shows the demographic characteristics of the 2 groups (i.e., ABOi and ABOc). Significantly more number of patients in ABOi group had chronic glomerulonephritis as the native kidney disease. Other demographic characteristics were comparable between the groups. One of the ABOi group patients was second transplant recipient. All the transplants in ABOi group were live related ones while in ABOc group 0.8% (n = 6) were deceased donor transplants.

Table 1 Demographic characteristics and clinical profile of patients

Table 2 shows the donor and recipient blood group distribution, starting antibody-titer and number of DFPP sessions required. It also shows the allograft biopsy details. The majority were O-blood group recipients (50%) followed by A and B (25% each). Most frequent titer was 1:256 (32%).

Table 2 Blood group and titer distribution in ABOi group

Recipient outcome is presented in Table 3. Mean duration of follow-up was 10.15 ± 9.34 months and 16.67 ± 9.63 months in ABOi and ABOc groups respectively. Patient and death censored graft survival was comparable between groups (P = 1). Figures 1 and 2 show Kaplan–Meier curves comparing patient and death censored graft survival between the two groups. Serum creatinine values at discharge and after 1-month were also comparable (P = 0.23 and 0.24 respectively). One patient died due to acute coronary syndrome. Although he had longstanding history of diabetes and hypertension, his pretransplant cardiac evaluation was normal. Another patient developed reduced urine output on the day of transplant. Graft biopsy showed thrombotic microangiopathy. There were no neutrophils or mononuclear cells in peritubular capillaries or glomeruli; neither there was any acute tubular injury. Staining for C4d in peritubular capillaries was negative. Repeat CDC and flow cytometry crossmatch was negative. Thrombotic microangiopathy was thought to be tacrolimus induced and hence it was withdrawn. He received plasma exchange sessions. He developed sepsis, which responded well to IV antibiotics. His renal functions and overall clinical condition started improving. Four weeks posttransplant, patient developed severe epigastric pain and recurrent bilious vomiting. Upper gastro-intestinal endoscopy showed hemorrhagic and necrotic ulcerative lesions in esophagus and stomach. Biopsy from these lesions revealed mucormycosis. Also, there was evidence of intranuclear inclusion bodies in gastric mucosa suggestive of CMV gastritis. He was treated with intravenous liposomal amphotericin B (3 mg/kg/day) and intravenous ganciclovir (2.5 mg/kg/day). Later his abdominal pain worsened, and he developed refractory hypotension. Exploratory laparotomy was done which showed 3 cm × 3 cm rent in posterior wall of stomach. Distal gastrectomy along with debridement and feeding jejunostomy was done. But despite these measures, he succumbed to sepsis.

Table 3 Recipient outcome

Close

Figure 1

Kaplan–Meier graph comparing patient survival between ABOi and ABOc group. (ABOi: ABO incompatible; ABOc: ABO compatible)

Export to PPT

Close

Figure 2

Kaplan–Meier graph comparing death censored graft survival between ABOi and ABOc group. (ABOi: ABO incompatible; ABOc: ABO compatible)

Export to PPT

On analysis, infection rates were not significantly different between the ABOi and ABOc groups (P = 0.16). BKV infection and pneumonia were seen in one patient each. As mentioned above, one patient had CMV infection and gastric mucormycosis.

A total of nine protocol and eight indication biopsies were done. Details of these biopsies are shown in Table 2. All protocol biopsies were normal. Staining for C4d was positive in 53% of cases. Of 8 patients whose biopsy was done for indication, two had delayed graft function. Of these, one had thrombotic microangiopathy as described above while another had diffuse cortical necrosis secondary to graft renal vein thrombosis. One patient had slow decline of serum creatinine in posttransplant period. In him, first renal biopsy showed acute tubular necrosis and protocol biopsy after 3 months showed borderline cellular rejection, which did not require any treatment, as renal function was stable. One of the patients developed nephrotic range proteinuria and active urine sediments. Graft renal biopsy showed C3 glomerulopathy. Her native kidney disease pretransplant was unknown. Her graft function is stable. Remaining 3 patients were biopsied for graft dysfunction. All had acute cellular rejections, which responded well to methylprednisolone pulse.

Discussion

In 1950s and 60s, initial attempts to do ABOi transplant in USA were met with high failure rate and very poor graft survival. It was concluded that ABO compatibility is a necessary prerequisite for successful renal transplant.[91011] In 1987, Alexandre et al. showed that successful ABOi transplants with good graft outcome could be achieved using pretransplant desensitization (or preconditioning) protocol. He used plasma exchange to remove anti-A or anti-B antibodies and splenectomy to prevent further antibody production. Antilymphocyte globulin was used for induction.[12] Subsequently, large number of ABOi transplants was done in Japan using preconditioning protocol of plasma exchange and splenectomy with good graft and patient outcome.[13] Later, it was realized that chemical splenectomy using anti-CD20 monoclonal antibody-Rituximab can replace surgical splenectomy with good results[14] and splenectomy associated complications could be avoided. Most centers have now stopped doing splenectomy for ABOi transplant. The majority of the European centers have used immunoadsorption, IVIG and rituximab based protocols with good success.[15161718] In United States, Montgomery et al. have used rituximab, plasma exchange, and CMV IVIG as preconditioning.[19] Selective methods of antibody removal such as immunoadsorption and DFPP is being preferred these days as large plasma volumes can be processed, and risk of bleeding due to elimination of coagulation factors is minimized. Lately, in an effort to reduce overall immunosuppression, Flint et al. and Montgomery et al. have done ABOi renal transplants without rituximab in their protocols with good graft and patient survival outcome.[2021] In our center, we have used preconditioning protocol of rituximab and conventional plasma exchange (5 patients)/DFPP (12 patients) followed by IVIG.

Patient survival was 90% in the current study. This was 98% and 100% in the study by Tydén et al., Flint et al. and Lipshutz et al. respectively.[162022] Death censored graft survival of ABOi transplant group in our study was 95%. This is similar to the graft survival in the study by Tydén et al. (97%), Flint et al.(100%), and Lipshutz et al.(94.4%).[162022] In the present study, death censored graft survival and patient survival of ABOi recipients were comparable to that of ABOc group. These findings are similar to that of study by Genberg et al.[15] In our study, serum creatinine at 1-month was similar between the 2 groups without any significant difference.

Rate of biopsy proven acute rejection was comparable between the groups. Biopsy proven acute rejection rate in ABOi recipients in the current study was 15%. This was better than the BPAR of 40% in a study by Wilpert et al. with similar immunosuppression protocol.[23] Lipshutz et al. reported BPAR of 11% in their ABOi recipients while it was 32% in a study by Uchida et al.[2224] Most of the studies have reported higher incidence of antibody-mediated rejection in ABOi recipients, which has varied from 5 to 33%.[22232425] In our study, none of the ABOi recipients developed antibody-mediated rejection. Although one patient had TMA on biopsy, other features of AMR such as neutrophils or mononuclear cells in peritubular capillaries or glomeruli, acute tubular injury and C4d deposition in peritubular capillaries, were missing.[26] In addition, his renal function improved after stopping tacrolimus, which supported it to be drug induced TMA.

Graft biopsies showed C4d staining in 53% of cases. In various studies, C4d positivity in protocol graft biopsies has been seen in 80–94% of patients and does not necessarily indicate antibody-mediated rejection.[2728] In fact, it may represent the phenomenon of accommodation, in which the graft continues to function normally despite the presence of anti-blood group antibodies.[28]

The infection rate in ABOi recipients was 15%, which was comparable to ABOc group. One ABOi patient developed CMV infection, and one had BKV infection. Most studies have shown a trend toward higher infection rate amongst ABOi transplant recipients, varying from 18 to 50%.[16172023]

There are some limitations of our study: The sample size of ABOi group is small. Also, the immunosuppression protocol was variable for ABOc group as mentioned in materials and methods. Follow-up of patients is short, and hence long-term outcome cannot be commented upon.

Concluding, published reports of ABOi transplant experience from developing countries are very few. Ours is one of the first such series. Graft and patient survival have been excellent in the current study. Rate of biopsy proven acute rejection was at par with ABO compatible ones and more importantly there were no antibody-mediated rejections. The incidence of posttransplant infections, which is a major concern in developing world, was acceptable and, in fact, lower than other similar studies. Our experience has been encouraging so far and proves that it is the right time for widespread implementation of ABOi transplants even in developing countries.