EXPERIMENTAL AND CLINICAL TRANSPLANTATION

We report a case of early recurrent membranous glomerulonephritis after kidney transplant from a deceased donor. The patient received induction therapy and was discharged with a serum creatinine level of 0.78 mg/dL on triple maintenance immunosuppressive therapy, which included tacrolimus, mycophenolate mofetil, and prednisone. At 7 months after transplant, a graft biopsy for new-onset isolated proteinuria (2.7 g/day) revealed stage 2 recurrent membranous glomerulonephritis. In the face of persistent proteinuria despite combined conservative rituximab therapy over several months and the total eradication of CD20-positive cells, bortezomib was introduced. This resulted in a substantial decline in proteinuria within 2 months and its subsequent disappearance several months later. This was paralleled by a considerable drop in plasma CD34-positive and CD138-positive cell counts. These preliminary observations indicate that recurrent posttransplant membranous glomerulonephritis is associated in part with a B-cell-mediated immunologic process that may involve both CD20-positive and plasma cells. Rituximab-resistant or partially responsive recurrent posttransplant membranous glomerulonephritis may benefit from a proteasome inhibitor-based therapy.

Key words : Autoimmune disease, Plasma cell, Proteasome inhibitors, Proteinuria, Recurrent posttransplant glomerulonephritis

A 50-year-old male patient presented with end-stage renal disease secondary to membranous glomerulonephritis (MGN). He underwent deceased-donor kidney transplant using induction therapy with rabbit antithymocyte globulin (ATG Fresenius, Grafalon; Neovii, Waltham, MA, USA), followed by a triple maintenance immunosuppressive regimen that included tacrolimus (Prograf; Astellas Pharma, Tokyo, Japan), mycophenolate mofetil (Cellcept; Hoffmann-La Roche, Basel, Switzerland), and prednisone. The patient was discharged with a serum creatinine level of 0.72 mg/dL. Tacrolimus dosage was adjusted according to whole blood trough levels. Mycophenolate mofetil was maintained at 2 g/day, and prednisone was tapered and maintained at 5 mg/day.

Overt proteinuria was first detected at 5 months posttransplant (2.2 g/day). Ramipril (Tritace; Sanofi-Aventis, Paris, France) was started at a dose of 10 mg/day. In the face of persistent isolated proteinuria with a negative cross-match and panel reactive antibodies, a graft biopsy was performed, which revealed characteristic features of stage 2 recurrent posttransplant MGN (PTMGN) (Figure 1). Irbesartan (Aprovel; Sanofi-Aventis) and subsequently spironolactone (Aldactone; Konoshima Chemical, Osaka, Japan) were added and increased to optimally tolerated dose levels. Proteinuria worsened to reach 8.5 g/day despite optimal triple renin angiotensin aldosterone system blockade, tacrolimus at a therapeutic whole blood trough level, and adequate mycophenolate mofetil dosage, which allowed systolic and diastolic arterial blood pressure levels to be maintained at 100 to 120 mm Hg and 70 to 80 mm Hg (Figure 2). Rituximab (Mabthera; Roche, Basel, Switzerland) was introduced in four 500-mg doses 1 month apart. Partial remission in proteinuria was noted, followed by a rebound that occurred after the third dose.

In the face of persistent proteinuria despite near eradication of CD20-positive cells (Figure 3), bortezomib (Velcade; Takeda Oncology, Cambridge, MA, USA), a proteasome inhibitor, was started at a dose of 1.3 mg/m for a total of 4 doses over 2 weeks. Within 2 months, considerable drops in CD34-positive and CD138-positive cell counts were noted that were paralleled by significant reductions in proteinuria, which resolved several months later (Figure 3).

Membranous glomerulonephritis may recur or develop de novo after transplant. The de novo form is twice as common as the recurrent form.¹ Recurrent PTMGN can occur in approximately 40% of patients, usually within the first year.² In our case, the overt proteinuria appeared at 7 months (Figure 2). Posttransplant MGN is characterized by a subepithelial immune complex dense deposit with isolated linear C4d deposition in the glomerular capillary basement membrane, as in our case, leading to predominantly epithelial cell damage without significant endothelial cell injury or inflammatory cell infiltration (Figure 1). The presence of multilayering of the peritubular capillary basement membrane with C4d deposition indicates its association with antibody-mediated injury³ and may represent another manifestation of chronic humoral rejection.⁴

Two types of antibodies are recognized: antibodies against the major human histocompatibility HLA complex and antibodies against a non-HLA major histocompatibility complex class I-related chain gene A (MICA) and gene B.⁵ Unlike classical HLA molecules, MICA expression is limited to dendritic cells, monocytes, activated T cells, keratinocytes, fibroblasts, and endothelial and epithelial cells. Products of the gene, MICA molecules, act as a ligand for several immune cells. A sizeable proportion of kidney transplant recipients may have anti-MICA antibodies present in their sera at the time of transplant that may be directed against their own MICA antigens, and many develop de novo MICA antibodies after transplant that may be either donor or nondonor specific.⁶ Antibodies directed against MICA may occur as a consequence of allogeneic donor sensitization and from some novel autoreactive mechanism yet to be identified, such as autoimmune disease.⁶

Analysis of MICA is neither easily nor routinely evaluated in most transplant centers. In our case, MICA antibodies were detected at the time of graft biopsy. Their donor specificity could not be identified and their de novo occurrence could not be confirmed. Given that MICA expression occurs on epithelial cells and that the predominant epithelial cell injury (podocyte) observed in our patient was in the absence of any alloimmune response, one can postulate a possible role of anti-MICA antibodies in the pathogenesis of PTMGN, irrespective of the timing of their occurrence. In the case of recurrence, they may represent preformed anti-MICA antibodies that might have been potentially responsible for the native kidney disease and may cause early recurrence posttransplant.

Phospholipase A2 receptor (PLA2R), a podocyte structure, has been recognized as an important specific marker for idiopathic MGN and PTMGN.^1,7 Persistence or reappearance of posttransplant anti-PLA2R is associated with increasing proteinuria and resistant disease.^8,9 Recent reports have described the presence of antibodies against MICA in kidney allografts localized in podocytes within the glomeruli of renal transplant recipients without or with acute rejection, together with infiltrating mononuclear cells, B cells, CD8-positive T cells, and natural killer cells.¹⁰ Whether PLA2R is a product of the MICA gene or the MICA molecule family remains to be determined. Unfortunately, staining and testing for PLA2R and the specific antibodies were not available at our center.

Induction therapy and standard maintenance immunosuppression in some cases are sufficient to induce immunologic remission; however, they do not preclude the possibility of recurrence of PTMGN,7 as was the case in our patient. Interestingly, in our patient, the increasing proteinuria was paralleled by a progressive rise in total lymphocyte count (TLC) while on the same maintenance immunosuppressive regimen (Figure 3). Surprisingly, the peak TLC coincided with the highest level of proteinuria in the absence of any rejection process, which was followed by partial simultaneous reduction in both TLC and proteinuria after induction therapy with rituximab (Figure 3). The considerable and persistent increase in TLC despite rituximab treatment and in the absence of any histologic or serologic evidence of any type of rejection may be explained, although on a purely speculative basis, by the proliferation of yet to be identified immune cell, such as memory B cells that may be involved in the pathogenesis of PTMGN. Alternatively, the significant increase in TLC may be related to a process of immunotolerance.¹¹ Unfortunately, none of the markers associated with immunotolerance or microchimerism have thus far been identified. The partial response to rituximab in our patient indicates, at least in part, a pathogenic role of CD20-positive B cells in PTMGN, as also reported previously.¹² Rituximab binds to the CD20 receptor of B cells and depletes them through cellular-dependent cytotoxicity and stimulation of B-cell apoptosis.¹³ It may also target short-lived plasma cells still expressing CD20.¹⁴ CD20 is expressed later in the B-cell lineage than CD19 and is not found on pro-B cells of mature plasma cells known to produce 90% of circulating immunoglobulin G.¹⁴

In our patient, CD20-positive cell counts decreased after the first dose of rituximab and then increased again. Rituximab eliminates peripheral B cells without preventing the regeneration of B cells from precursor cells and does not directly affect immunoglobulin G levels.¹⁴ This may explain at least in part the reappearance of proteinuria during treatment with rituximab. The partial remission observed with anti-CD20 therapy and the relapse in proteinuria despite near elimination of CD20-positive cells suggests a possible additional CD20-independent mechanism of antibody production. The possible role of memory plasma cell in the persistent production of alloantibodies or autoantibodies has also been suggested but so far remains unproven.¹¹ Classically, memory B cells can become activated plasmablasts, which then convert to plasmocytes within hours of exposure to an antigen. Plasma cells produce large quantities of protein immunoglobulin G. Inhibition of proteasome activity results in accumulation of misfolded proteins and consequent apoptosis of plasma cells.¹⁴

In vitro, bortezomib induces apoptosis of all antibody-producing plasma cells.¹⁵ In the absence of rituximab-induced complete remission despite near eradication of a CD20-positive cell count, bortezomib was introduced. Within 2 months, considerable drops in CD34-positive and CD138-positive cell counts were noted in our patient, which was paralleled by a significant reduction in proteinuria that disappeared several months later (Figure 3). Both CD34 and CD138 are established plasmocyte markers.

Our clinical observation suggests a possible role of plasma cells in the pathogenesis of recurrent PTMGN, mainly in those patients who are resistant to CD20-positive targeted therapy. To our knowledge, this is the first case report describing the successful use of a proteasome inhibitor (bortezomib) in recurrent PTMGN with persistent proteinuria that was refractory to combined conventional rituximab therapy.

Volume: 15 Issue: 3 June 2017

FULL TEXT