Amyloidosis is a generic term for a group of diseases

Transcription

1 Typing of Amyloidosis in Renal Biopsies Diagnostic Pitfalls Anjali A. Satoskar, MD; Kelly Burdge, MD; Daniel J. Cowden, MD; Gyongyi M. Nadasdy, MD; Lee A. Hebert, MD; Tibor Nadasdy, MD Context. Amyloidosis represents a group of diseases with extracellular deposition of congophilic fibrils of similar morphology but differing chemical composition. The types commonly involving the kidney are AL (light chain amyloid) and AA (serum amyloid A). Familial amyloidosis can also affect the kidney, but we have not encountered such a case during the study period. Distinguishing between the AL and AA forms of amyloid is clinically important because of the different treatments and outcomes. The classification of amyloidosis is made by immunostaining with antibodies to and immunoglobulin light chains and for serum amyloid A protein. Objective. To draw attention to the nonspecific immunofluorescence staining patterns in renal biopsies with amyloidosis, causing potential diagnostic pitfalls. Design. Renal biopsies from 15 patients, including 13 cases of AL and 2 cases of AA amyloidosis, were studied. Immunofluorescence staining with routine antibody panel and immunoperoxidase staining for amyloid A were performed. Results. Of the 13 cases of AL amyloidosis, 2 cases showed little difference in staining intensity between and light chains (2 and 3, respectively) and 4 cases showed only moderate intensity (2 ) of the predominant light chain. The 2 cases diagnosed as AA amyloidosis also exhibited staining for light chains. One case had strong (3 ) signal for and moderate (2 ) for light chain, while the other showed weaker staining. Conclusions. Immunofluorescence staining for immunoglobin light chains on renal biopsy, as the first step to differentiate between AL and AA amyloidosis, may sometimes be inconclusive or even misleading. Applying amyloid A immunostain on a routine basis and detailed clinical history are essential to avoid misclassification. (Arch Pathol Lab Med. 2007;131: ) Amyloidosis is a generic term for a group of diseases that have in common tissue deposits of extracellular fibrillar proteins of similar ultrastructure, different chemical composition, and different etiologies. Amyloidosis is a relatively uncommon cause of renal disease and proteinuria, 1 3 but the incidence increases with age. In patients older than 60 years, amyloidosis is found in 10% to 12% of patients with nephrotic range proteinuria. 4,5 At our institution, during the period of this study, 2% of all adult native renal biopsies were diagnosed as amyloidosis. There are at least 21 different proteins that are capable of forming amyloid fibrils. 6 The forms that are frequently associated with renal involvement, and commonly encountered in renal pathology practice, are AL and AA amyloidosis. Familial/hereditary form of amyloidosis can also affect the kidney but are rarely encountered in a routine diagnostic renal pathology laboratory. In the United States, the overwhelming majority of renal amyloidosis cases are the AL (amyloid light chain) type. The second most common form is AA amyloidosis derived from se- Accepted for publication November 15, From the Departments of Pathology (Drs Satoskar, Cowden, G. M. Nadasdy, and T. Nadasdy) and Internal Medicine (Drs Burdge and Hebert), The Ohio State University Medical Center, Columbus. The authors have no relevant financial interest in the products or companies described in this article. Reprints: Tibor Nadasdy, MD, Department of Pathology, The Ohio State University, Starling Loving Hall M018, 320 W 10th St, Columbus, OH ( Nadasdy-1@medctr.osu.edu). rum amyloid A protein, an acute phase reactant synthesized in the liver, seen in conditions associated with chronic immune activation such as chronic infections (tuberculosis, osteomyelitis, bronchiectasis), chronic inflammatory diseases such as rheumatoid arthritis, ankylosing spondylitis, Crohn disease, or familial Mediterranean fever. For editorial comment, see p 850. All forms of amyloid share 4 common properties: (1) homogeneous, amorphous, and faintly eosinophilic extracellular deposits in hematoxylin-eosin stained sections; (2) Congo red positivity with apple green birefringence under polarized light; (3) ultrastructurally, randomly oriented fibrils measuring 7 to 10 nm in diameter; and (4) a -pleated sheet conformation, as seen by x-ray crystallography. The first 3 properties are used for routine histopathologic diagnosis; however, these methods cannot differentiate between the amyloid subtypes. Differentiating between them is important because their pathogenesis and treatment strategies differ. Because AL amyloidosis is the most common type encountered in the United States, and it is known to contain monoclonal immunoglobulin light chains, the first step is immunofluorescence (IF) staining for and light chains. A clear predominance and bright staining of one light-chain type compared with the other supports the diagnosis of AL amyloidosis. Absence of Arch Pathol Lab Med Vol 131, June 2007 Typing of Renal Amyloidosis Satoskar et al 917

2 Figure 1. Average immunofluorescence staining intensities in the subytpes of amyloid for each of the 9 antibodies, albumin (Alb), immunoglobulin (Ig) G, IgA, IgM, C1q, C3, fibrinogen (Fib), light chain, and light chain, respectively. staining for immunoglobulin light chains in the amyloid deposits in a renal biopsy is suggestive of AA amyloidosis. This is confirmed by immunostaining for serum amyloid A protein. However, in our experience, the clearcut differentiation between AL and AA amyloid, as described previously, may occasionally be difficult due to inconclusive staining patterns with immunoglobulin light chains. Inconclusive staining may also be associated with hereditary/familial subtypes of amyloid. 7 In this study, we would like to draw attention to potential pitfalls in the characterization of the most common amyloid subtypes seen in renal biopsy specimens, AL and AA. MATERIALS AND METHODS Biopsy Material We identified renal biopsies from 17 patients with renal amyloidosis diagnosed between January 1, 2003, and August 1, 2005, at our institution. Two cases were excluded from the study due to lack of adequate tissue for accurate subtyping. As part of the routine diagnostic protocol, each biopsy specimen was divided for light microscopy, ultrastructural examination, and direct IF staining with a panel of antibodies to albumin, immunoglobulin (Ig) G, IgA, IgM, C1q, C3, fibrinogen, and and light chains. Paraffin-embedded sections were stained with hematoxylin-eosin, periodic acid Schiff, Jones methenamine silver, and Masson trichrome. Congo red stain and amyloid A immunoperoxidase stains were done on all the biopsies. IF Staining Tissue was embedded in optimal cutting temperature (OCT) compound (Tissue-Tek, Torrance, Calif) and snap frozen in liquid nitrogen. Frozen sections were cut at 3- m thickness and fixed in cold acetone (4 C) for 10 minutes. Direct IF staining was performed using fluorescein isothiocyanate conjugated primary polyclonal rabbit anti-human antibodies to IgG, IgA, IgM, C1q, C3, fibrinogen, and and immunoglobulin light chains (all from Dako, Carpinteria, Calif). The slides were examined under a Zeiss Axioscop IF microscope and the intensity of staining was graded on a scale from 0 to 3 (1, mild staining; 2, moderate staining; 3, to strong staining; / indicated trace staining). If we had difficulty categorizing the staining intensities into 1, 2, or3, wegradeditas1 to 2 (scored as 1.5) or 2 to 3 (scored as 2.5). Images were captured using a Carl Zeiss 3CD digital camera. Electron Microscopy Tissue was placed in vials containing 3% glutaraldehyde (ph 7.4) and postfixed in osmium tetroxide. Following routine processing and embedding into Spurr resin, thin sections were cut and contrasted with uranyl acetate and lead citrate. Ultrastructural examination was performed using Carl Zeiss EM 900 transmission electron microscope. Indirect Immunoperoxidase Staining for Serum Amyloid A Protein Sections were deparaffinized and rehydrated. Antigen retrieval was followed by immunostaining performed in an autostainer (Dako LV-1 Autostainer) using the Dako Universal staining system. Sections were stained with primary polyclonal rabbit antihuman amyloid A antibody (Alpco Diagnostic, Windham, NH) followed by biotinylated goat anti-rabbit secondary antibody (Vector Laboratories Inc, Burlingame, Calif). The Vectastain Elite detection kit (Vector) was used with 3,3 -diaminobenzidine chromogen as the detection system. Slides were counterstained with hematoxylin (Richard-Allen Scientific, Kalamazoo, Mich). RESULTS Of the 15 total cases included in the study, each of the 15 patients had proteinuria at the time of diagnosis; 13 of them had nephrotic range proteinuria. Of the 15 total patients, 13 (86%) were diagnosed as AL amyloidosis, and 2 (13.3%) were diagnosed as amyloidosis AA. Average IF staining intensities in the subtypes of amyloid for each of the 9 antibodies, albumin, IgG, IgA, IgM, C1q, C3, fibrinogen, light chain, and light chain, respectively, are shown in Figure 1. Biopsies With AL Amyloidosis Of the 13 patients with AL amyloidosis (patients 1 through 13 in the Table), 10 patients had chain amyloidosis, and 3 had chain amyloidosis. Nine of the 13 biopsies exhibited the expected pattern of bright amyloid staining for the pathogenic light chain and weak or no staining for the other light chain (Figure 2, A and B). Two biopsies (patients 3 and 5) showed little difference in staining intensity between (3 ) and (2 ) light chains (Figure 3, A and B). If we arbitrarily define a striking difference in intensity as more than or equal to 1.5, these 2 cases did not show striking differences between and staining. Patient 5 did have an IgG spike in the serum and 918 Arch Pathol Lab Med Vol 131, June 2007 Typing of Renal Amyloidosis Satoskar et al

3 Patient No. Age, y Clinical and Immunostaining Characteristics* Amyloid Subtype IF IF AA IHC Clinical History 1 58 AL 0 2 Neg Transplant kidney with recurrent amyloidosis; has recurrent clonal plasmacytosis 2 78 AL Neg Had urine monoclonal spike 3 77 AL 2 3 Neg No history of monoclonal spike, patient developed melanoma; no further treatment for amyloidosis 4 72 AL 0 2 Neg No monoclonal spike, but patient died soon after the biopsy 5 78 AL 2 3 Neg IgG spike in serum and urine; but patient very ill with severe ischemic heart disease; he was not a candidate for aggressive intervention 6 76 AL 1 3 Neg No monoclonal spike; had severe coronary artery disease; developed mesenteric vascular ischemic injury, suspected viscus perforation; no further treatment for amyloidosis 7 44 AL Neg IgG spike in serum and urine 8 73 AL 1 3 Neg IgG spike in serum and urine; bone marrow biopsy revealed monoclonal plasma cell population 9 48 AL Neg Bone marrow biopsy had light chain restricted plasma cell dyscrasia AL 0 3 Neg No history of monoclonal spike; bone marrow biopsy revealed plasmacytosis but not sufficient for myeloma; was treated with chemotherapy; patient required dialysis AL Neg IgG spike in serum and urine; cardiac and vascular involvement; bone marrow biopsy was not consistent with myeloma; patient treated with prednisone with plan for chemotherapy AL 2 0 Neg Bone marrow biopsy revealed clonal B lymphocytes; also had amyloid angiopathy and peripheral nerve involvement AL Neg Biclonal IgG peak in serum and urine; bone marrow biopsy revealed 9% plasmacytosis; patient being treated with dexamethasone AA 3 2 Pos Wegener granulomatosis; no monoclonal spike in serum or urine AA Pos COPD, recurrent lung infections, and pleural effusions; faint IgM spike in serum * IF indicates immunofluorescence; AA, amyloid A; IHC, immunohistochemistry; AL, amyloid light chain; Neg, negative; Ig, immunoglobulin; Pos, positive; and COPD, chronic obstructive pulmonary disease. and immunofluorescence staining intensities semiquantitatively scored from 1 to 3 (see Materials and Methods ). AA immunostain showed nonspecific background staining in tubular epithelium and serum in glomerular capillary loops. urine, supporting the diagnosis of light-chain amyloidosis. The AA immunostain was negative. In patient 3, there was no history of monoclonal spike at the time of the biopsy, but the immunostain for AA protein was negative. Clinical follow-up revealed that patient 3 was treated with steroids and immunosuppressive agents for 4 months, because he also had crescentic glomerulonephritis. Following this, he was diagnosed with melanoma. Therefore, no specific treatment was given for his amyloidosis. Serum creatinine now is 1.8 mg/dl. If we define strong staining as 3, then 4 of the 13 patients (patients 1, 4, 11, and 12 in the Table) of AL amyloid had only moderate staining intensity (2 ). Patients 1, 11, and 12 did have a monoclonal spike. Patient 12 also had a bone marrow biopsy, which showed clonal B lymphocyte population. Patient 1 was treated with autologous bone marrow transplant for AL amyloidosis. Patient 4 did not have a monoclonal spike at the time of biopsy; AA immunostain was negative. Unfortunately, the patient died soon after the biopsy. Of the 13 patients with AL type amyloidosis, 8 had a known monoclonal protein spike in the serum or urine or both and 1 patient had a biclonal spike of IgG and, and in 4 of the patients there was absent monoclonal spike on testing (patients 3, 4, 6, and 10). In all these biopsies, the AA stain was negative. The biopsies of 3 of the cases (2, 12, and 13) had some staining for the amyloid A protein in the tubular epithelium and in the serum in glomerular capillaries, but the amyloid deposits themselves were negative. Biopsies With AA Amyloidosis Case 14 (Table). The biopsy of patient 14 showed bright staining (3 ) for light chain and moderate (2 ) for light chain (Figure 4, A and B). The patient did not have monoclonal spike in the serum or urine. Congo red stain was positive with apple green birefringence under polarized light (Figure 4, D). Immunostain for AA protein was strongly positive (Figure 4, C). The Congo red positivity (Figure 5, A) was eliminated by potassium permanganate pretreatment (Figure 5, B). Ultrastructural examination revealed typical amyloid fibrils. The patient had a long history of chronic nasal mucosal ulcers with episodes of epistaxis, saddle-nose deformity, and chronic sinusitis. Nasal mucosal biopsy revealed chronic inflammation and necrosis. Computed tomography scan showed nasal mucoperiosteal thickening. High C-reactive protein and elevated sedimentation rate were in favor of a chronic inflammatory disease process. Wegener granulomatosis was considered. Renal biopsy showed globally solidified glomeruli with (secondary) AA-type amyloid deposits in the glomeruli as well as in the vascular wall. Arch Pathol Lab Med Vol 131, June 2007 Typing of Renal Amyloidosis Satoskar et al 919

4 Figure 2. Direct immunofluorescence staining for and light chains in AL (amyloid light chain) amyloidosis. This case demonstrates the striking difference in staining intensities between the 2 immunoglobulin light chains and, depicting monoclonality. A, Negative/background staining for light chain (immunofluorescence, original magnification 400). B, 3 staining of amyloid deposits in the mesangium for light chain (immunofluorescence, original magnification 400). Figure 3. Case of AL (amyloid light chain) amyloidosis ( chain type). A, light-chain staining in a small artery with amyloid deposits (immunofluorescence, original magnification 200). B, Prominent light-chain staining in the same artery (immunofluorescence, original magnification 200). There is no striking difference in intensities despite the monoclonal nature of the amyloid. Case 15 (Table). This biopsy also showed staining for immunoglobulin light chains, albeit weaker (1, ; 0.5, ). To our surprise, this patient showed a faint band of restricted mobility in the region on serum electrophoresis with faint IgM on serum immunofixation. No monoclonal spike was found in the urine. Immunostain for serum AA protein was strongly positive. The patient had a history of chronic obstructive pulmonary disease, recurrent pulmonary infections, and pleural effusions. No overt plasma cell or B-cell lymphoproliferative disease was present. Bone marrow biopsy was not performed. Because of the small number of cases in each group, a statistical analysis has not been performed, but a general observed trend has been described. COMMENT Although amyloidosis is a relatively uncommon cause of renal disease and proteinuria, the renal pathologist is commonly involved in the diagnosis of amyloidosis. Familial/hereditary forms of amyloidosis are rare but occasionally may involve the kidney. As indicated in the introduction, clinically relevant renal amyloidosis is overwhelmingly caused by AL or AA amyloidosis. In this study, we focused on AL and AA amyloidosis. More recently, treatment with high-dose melphalan and autologous stem cell transplantation has led to improved outcomes in cases with AL amyloidosis. 8,9 Treatment of AA amyloidosis is targeted at controlling the underlying medical condition as well as at inhibiting fibril formation (eg, colchicine). 10,11 Because of the divergent treatment approaches, the differentiation of AL and AA amyloid in a renal biopsy is very important. This is usually not a problem if there is monoclonal spike in serum and urine, monoclonal plasmacytosis in a bone marrow biopsy, and bright immunostaining for the corresponding light chain on renal biopsy tissue. However, sometimes there may be no history of monoclonal spike, and the immunostaining on biopsy may be inconclusive. The main pitfalls that we have encountered are (1) we do not always see a striking predominance of one light chain over the other in AL am- 920 Arch Pathol Lab Med Vol 131, June 2007 Typing of Renal Amyloidosis Satoskar et al

5 Figure 4. Case of AA (amyloid A) amyloidosis with nonspecific light-chain staining. A, 3 light-chain staining (immunofluorescence, original magnification 100). B, 2 light-chain staining in the glomerular amyloid deposits (immunofluorescence, original magnification 100). C, Indirect immunoperoxidase staining of paraffin-embedded, formalin-fixed tissue for serum AA protein shows positive staining of amyloid deposits (immunoperoxidase, original magnification 200). D, Apple green birefringence under polarized light of Congo red stained formalin-fixed tissue (original magnification 400). Figure 5. Case of AA (amyloid A) amyloidosis (depicted in Figure 4). A, Positive Congo red staining of the amyloid deposits in the mesangium and arterial walls (original magnification 200). B, Pretreatment with potassium permanganate eliminates Congo red staining proving the AA nature of the amyloid (original magnification 200). Arch Pathol Lab Med Vol 131, June 2007 Typing of Renal Amyloidosis Satoskar et al 921

6 yloidosis, (2) the intensity of staining is not always strong for light chains in AL amyloidosis, and (3) there is occasional nonspecific light-chain staining seen in AA amyloidosis. We hypothesize 2 possibilities for the nonspecific staining for immunoglobulin light chains in these cases of AA amyloid: (1) an in vitro phenomenon due to increased affinity to amyloid AA protein by the antibodies used or (2) an in vivo innate stickiness of the amyloid AA protein to immunoglobulin components in the serum. This sticky physical property is known for renal biopsies of diabetic patients, which frequently show intense staining of the glomerular capillary basement membranes, mesangium, and tubular basement membrane for albumin, IgG, and the light chains. However, in cases of AA amyloidosis, staining for albumin was not increased; in fact, both cases in this study had negative staining for albumin, but staining for IgG and IgA was present (Figure 1). Perhaps this is because of anionic charges in albumin and amyloid repelling each other, but this is only speculative. Therefore, it does not exhibit a generalized increased stickiness but just a selective affinity to immunoglobulin components. The lack of striking difference in staining intensity between and light chains in some cases of AL amyloidosis is difficult to explain. Occasional lack of strong intensity of staining for the predominant light chain may be due to several reasons; antigen masking could be one of them. Another reason could be that the main constituent of amyloid deposits in some cases of AL amyloid is the light-chain variable region alone (without parts of the constant region or of the whole immunoglobulin). 1 Thus, in such cases, the routinely used antibodies to the light-chain constant region would be ineffective in staining the amyloid deposits despite their AL nature. It is unlikely that technical issues with staining could account for this. Nine of the 13 cases of AL amyloidosis did show strong staining. We titrate every new batch of the antibodies using a positive control tissue. Also, every new vial of antibody in each batch is tested on the positive control tissue. Staining was done by the same technologist, with extensive experience, using the same protocol. Novak et al 12 have also reported negative or equivocal IF staining for and light chains in 12 of 36 renal biopsies with AL amyloidosis. All 12 cases were subsequently found to have a plasma cell dyscrasia with monoclonal immunoglobulin and/or free light chains on electrophoresis and immunofixation of urine or serum. Therefore, one has to keep in mind that inconclusive immunostains for light chains do not always exclude the possibility of AL amyloidosis. These cases are potential pitfalls in the accurate subclassification of renal amyloidosis. Recently, a group from the United Kingdom reported a series of 350 patients, in whom a diagnosis of AL amyloidosis was suggested based on clinical history, biopsy findings, and absence of family history of amyloidosis. 7 Of these 350 patients, 34 were subsequently found to have amyloidogenic mutations, in genes encoding fibrinogen A -chain and transthyretin. In all 34 cases, the diagnosis of hereditary amyloidosis was later confirmed. In 8 (24%) of these 34 patients, intact monoclonal immunoglobulins were detected in the serum (at levels 0.2 g/dl) but no free light chains were detected in the urine. Thus, just a small monoclonal spike in the serum does not necessarily prove that the amyloidosis is light-chain related, because monoclonal gammopathy of undetermined significance is not a very rare finding. On the other hand, in approximately 10% of patients with AL amyloidosis, serum and/ or urine electrophoresis studies may be negative. 1 To add to the complexity, in the series of Lachmann et al 7 there was no specific immunohistochemical staining in the amyloid deposits with antibodies to or light chains in 195 of 316 patients with AL amyloidosis. However, circulating paraproteins or urinary free light chains were present in the latter. This study also emphasizes the fact that inconclusive immunostains for light chains in renal biopsies do not always exclude the possibility of AL amyloidosis. In our series, 3 biopsies (patients 2, 12, and 13) did show background immunostaining with the antibody to serum AA protein in tubular epithelial cells and serum accumulated in glomerular and peritubular capillary lumina. This can be misleading. It is important to interpret the AA immunostain in correlation with the Congo red stain for the localization of the amyloid deposits. In fact, all immunostains for amyloid should be interpreted in correlation with the Congo red stain. CONCLUSION In summary, unequivocal identification of the deposited amyloidogenic protein in renal and other tissue biopsies is essential to initiate appropriate treatment. Unfortunately, the immunostaining pattern for AL and AA amyloid (the most common subtypes found in renal biopsies) may sometimes be inconclusive; hence differentiation of amyloid types on renal biopsy may be difficult. Our recommendation is that the AA immunostain should be performed routinely, unless there is strong and selective staining in the amyloid deposits for the same light chain that is found in the monoclonal spike in the serum and urine. If still in doubt, consultation should be obtained from a center where more extensive and detailed immunostaining for the other amyloidogenic proteins can be performed. The diagnostic approach is multidisciplinary and demands careful clinical evaluation and histochemical and immunohistochemical studies. References 1. Falk RH, Comenzo RL, Skinner M. The systemic amyloidoses. N Engl J Med. 1997;337: Rivera F, Lopez-Gomez JM, Perez-Garcia R, et al. Frequency of renal pathology in Spain Nephrol Dial Transplant. 2002;17: Schena FP, the Italian Group of Renal Immunopathology. Survey of the Italian registry of renal biopsies, frequency of the renal diseases for 7 consecutive years. Nephrol Dial Transplant. 1997;12: Haas M, Meehan SM, Karrison TG, et al. Changing etiologies of unexplained adult nephrotic syndrome: a comparison of renal biopsy findings from and Am J Kidney Dis. 1997;30: Cameron JS. Nephrotic syndrome in the elderly. Semin Nephrol. 1996;16: Merlini G, Bellotte V. Molecular mechanisms of amyloidosis. N Engl J Med. 2003;349: Lachmann HJ, Booth DR, Booth SE, et al. Misdiagnosis of hereditary amyloidosis as AL (primary) amyloidosis. N Engl J Med. 2002;346: Dember LM, Sanchorawala V, Seldin DC, et al. Effect of dose-intensive intravenous melphalan and autologous blood stem cell transplantation on AL amyloidosis-associated renal disease. Ann Intern Med. 2001;134: Sanchorawala V, Wright DG, Seldin DC, et al. An overview of the use of high dose melphalan with autologous stem cell transplantation of AL amyloidosis. Bone Marrow Transplant. 2001;28: Zemer D, Pras M, Sohar E, Modan M, Cabili S, Gafni J. Colchicine in the prevention and treatment of the amyloidosis of familial Mediterranean fever. N Engl J Med. 1986;314: Okuda Y, Takasugi K. Successful use of a humanized anti-interleukin-6 receptor antibody, tocilizumab, to treat amyloid A amyloidosis complicating juvenile idiopathic arthritis. Arthritis Rheum. 2006;54: Novak L, Cook WJ, Herrera GA, Sanders PW. AL-amyloidosis is under diagnosed in renal biopsies. Nephrol Dial Transplant. 2004;19: Arch Pathol Lab Med Vol 131, June 2007 Typing of Renal Amyloidosis Satoskar et al