Drug-induced de novo thrombotic microangiopathy diagnosed 2 years after renal transplantation: a case report and literature review

Post-transplant de novo thrombotic microangiopathy (TMA) is a rare yet serious complication that generally can develop in renal transplant recipients immediately after reperfusion or several months after transplantation. Here, we report a case of systemic tacrolimus-associated TMA in a patient diagnosed 2 years after renal transplantation. A 49-year-old woman presented with severe anemia 18 months after undergoing renal transplantation. Anemia was refractory to recombinant human erythropoietin and was suspected to be due to excessive menstruation. Anemia persisted even after hysterectomy, and thereafter, pancytopenia developed. A bone marrow biopsy was performed and showed no evidence of myeloproliferative neoplasms. Furthermore, an increase in serum lactate dehydrogenase level and the appearance of schistocytes on peripheral blood smear was noted 24 months post-transplant. Other possible causes of de novo TMA were excluded, and an allograft biopsy was performed. Pathological findings of the allograft biopsy showed that some afferent arterioles had formed thrombi. Suspecting tacrolimus to be the cause of TMA, 25 months after the transplant, we switched treatment to cyclosporine. Pancytopenia and renal function improved after switching to this calcineurin inhibitor. Subsequently, her allograft renal function stabilized for three years after renal transplantation. We encountered a case of secondary drug-induced TMA in the late stages of renal transplantation. Therefore, TMA should be suspected when anemia with hemolysis is observed in recipients of kidney transplant.

Here, we report a case of tacrolimus (TAC)-associated TMA in a patient diagnosed 2 years after renal transplantation; successfully treated by replacing TAC with cyclosporine (CyA).

Case presentation
A 49-year-old woman with end-stage renal failure due to IgA nephropathy received a pre-emptive ABO-blood type-compatible renal transplant from her mother. In addition to nephrogenic anemia, the patient experienced anemic fluctuations due to excessive menstruation. Complement-dependent cytotoxic crossmatch and flow cytometric crossmatch tests yielded negative results. Anti-HLA antibody testing was weakly positive for both Class I and II, using HLA antigen-coated synthetic flow beads (Flow PRA Screening Kit, One Lambda Inc., Canoga Park, CA, USA). She also gave birth three times.
Since the patient was a highly sensitized case and she had received rituximab preoperatively. Induction immunosuppressive therapies included TAC, mycophenolate mofetil (MMF), methylprednisolone (mPSL), and basiliximab (BXM). Figure 1 shows the clinical course of the renal transplantation. Although improvement in post-operative renal function was good, hemoglobin remained at 6-8 mg/dL and was refractory to recombinant human erythropoietin (rHuEPO). Therefore, anemia was suspected to be caused by excessive menstruation. Since red blood cell transfusion was repeatedly required for 18 months post-transplantation, hysterectomy was performed at 20 months. Around the same time, serum creatinine levels increased from 1.45 to 2.17 mg/dL. Anemia persisted even after hysterectomy, with new clinical features, including episodes of nausea, fatigue, and abdominal pain. Furthermore, blood investigations revealed pancytopenia. Consequently, she underwent a bone marrow biopsy which revealed no myeloproliferative changes. MMF-induced myelosuppression was suspected; thus, treatment was switched to everolimus (EVR). However, no improvement in renal function or pancytopenia was seen, as anticipated. Thus, de novo TMA was suspected due to increased levels of serum lactate dehydrogenase (LDH) and the presence of schistocytes in the peripheral blood smear. The disintegrin and metalloprotease with thrombospondin type I repeats-13 (ADAMTS13) activity were within normal range (72%), and ADAMTS13 inhibitor was negative. C3 and C4 complement levels were all within normal range. No abnormalities were found in the coagulation or fibrinolytic blood profiles. Other possible causes of de novo TMA, such as bacterial or viral infections (Epstein-Barr virus, cytomegalovirus, and polyoma virus), were excluded. Doppler ultrasonography of the allograft revealed normal blood flow and no increased vascular resistance, following which, an allograft biopsy was performed. Pathologically, the allograft biopsy showed no signs of antibody-mediated rejection (C4d negative). And no afferent arterioles showed thrombus formation, which is not a typical histological finding of TMA (Fig. 2). Suspecting TAC as the cause for TMA, TAC was replaced with CyA 25 months after transplantation. The trough value of TAC was between 5-6 ng/mL.
We found that renal function and pancytopenia improved after switching from CNIs. Additionally, RBC fragmentation disappeared and the LDH level normalized. The patient's fatigue and abdominal pain reduced rapidly; renal function subsequently stabilized for three years after renal transplantation.

Discussion
TMA is a condition in which microvascular thrombosis occurs due to abnormalities in the vessel walls of small arteries and capillaries. TMA typically presents as a triad of hemolytic anemia, thrombocytopenia, and organ damage due to microcirculatory disturbances. Overall, de novo systemic TMA is reported to occur mostly within the first month after transplantation [10]. Post-transplant TMA can be diagnosed pathologically on a routine renal biopsy without evidence of laboratory abnormalities [11]. Table 1 shows a literature review of case reports of de novo TAC-induced TMA after renal transplantation [4,9,[11][12][13][14][15][16][17].
Diagnosis of hemolytic anemia is important for the early detection of TMA; thus, LDH levels, peripheral blood smear, Coombs test, haptoglobin, and other TMAoriented tests should be performed. In our case, although orthocytic anemia, pancytopenia, and renal impairment were observed, hemolytic changes such as elevated LDH and the appearance of schistocytes were not identified until 20 months post-transplantation. Thus, a delayed onset of symptoms during the prime period of TMA makes diagnosis difficult. Here, considering that two years had passed since renal transplantation, we had to rule out the following conditions: AMR, TMA development due to infection, drug-induced myelosuppression, and myeloproliferative neoplasm. Accordingly, we came to an exclusionary diagnosis of TMA due to TAC.
CNIs are known to cause transplant-related TMA, especially in the early period (3-6 months) after transplantation [8,9]. Various mechanisms contribute to the onset of TMA with the use of CNIs. Previous studies on rats treated with CyA reported that biosynthesis of vasoconstricting substances, such as thromboxane A2 and endothelin, increased; the expression of vasoconstrictor molecules, such as prostaglandin E2 and prostacyclin, decreased. This suggested that a loss of equilibrium in vasoactive peptides leads to arteriolar vasoconstriction [18]. Platelet-activating, pro-coagulant, and antifibrinolytic effects of CNIs have also been implicated in the development of TMA, especially in cases where the endothelium is already injured due to other mechanisms, such as ischemia-reperfusion injury and antibodymediated rejection [19,20]. The incidence rate of druginduced TMA due to TAC is less frequently reported than that due to CyA [9]. CyA-induced TMA has been reported to develop in a dose-dependent manner [21], consistent with the report that CyA is associated with vascular endothelial damage. However, the association between the onset of TAC-induced TMA and its dose remains unclear [9].
In a study of 26 patients who underwent renal or combined kidney-pancreas transplantation, TMA was  diagnosed by episodic allograft biopsy. In this study, 24 and 2 patients received CyA and TAC, respectively. Systemic TMA leads to early graft rejection more frequently than does localized TMA. Only 2 of the 26 patients receiving systemic TMA showed changes other than those seen in renal histology [22]. In a study by Schwimmer et al., 54% of patients receiving systemic TMA developed acute kidney injury requiring dialysis, and 38% showed TMA-related graft loss. Conversely, no patients with localized TMA required dialysis or experienced early graft loss [1].
Optimal treatment for TMA after renal transplantation, including discontinuation of the suspected drug and modification of immunosuppressive agents, remains controversial. Kwon et al. reported that changing the CNI resulted in good graft function at one year in 81% of TMA patients post-transplantation [23]. On the other hand, Zarifian et al. showed that although an improvement was observed in 81% of patients who switched to another CNI, 30.1% lost allograft renal function due to TMA [22]. Given that the cause of CNI-induced TMA is vascular endothelial damage, plasma exchange is often ineffective [22].
It has been reported that in cases of secondary TMA development (including infection, ischemia-reperfusion injury, and drugs), abnormal complement regulation may have occurred, and triggered the development of TMA [24]. Thus, withdrawal of CNI treatment alone might not be sufficient to reverse adverse effects of the alternate pathway of complement activation associated with these conditions [2,9]. In such cases, plasma exchange could be useful. Furthermore, eculizumab, an anti-complement C5 inhibitor, has shown efficacy in many resistant cases of de novo post-transplant TMA [25][26][27], including TAC-induced TMA [16]. Hence, the use of eculizumab can be considered in refractory cases after CNI switching.
One pediatric case of localized TAC-induced TMA showed complete resolution after switching to CNIs, with no chronic changes on repeat biopsy [11]. In another study including 26 patients with CyA-induced TMA, 11 patients underwent post-TMA biopsies that showed a high percentage of chronic vascular changes (73%) [23]. However, de novo TMA can recur after switching to CNIs [4]. Renal allograft survival is clearly reduced with the development of post-transplant TMA [23,28,29] and is worse with recurrent disease [30]. In our case, renal function progressively improved after switching to CNI, and there was no recurrence for one year; however, long-term follow-up of renal function proved crucial in preventing TMA recurrence.
TAC plays a central role in the prevention of graft rejection, but can be associated with adverse events, such as nephrotoxicity, and can lead to TMA. The diagnosis required several differentiations; however, the CNI change was effective, and renal function improved to pre-treatment levels. Long-term followup is required in the future.

Conclusion
We encountered a case of secondary drug-induced TMA after renal transplantation. We conclude that TMA should be suspected in cases presenting with anemia and hemolysis, regardless of the time since renal transplantation.