The incidence of thin glomerular basement membrane

Transcription

1 Approach to the iagnosis of Thin Basement Membrane Nephropathy in Females With the Use of Antibodies to Type IV Collagen Ginette Lajoie, M, FRCPC Context. Thin basement membrane nephropathy is recognized by a diffusely thin glomerular basement membrane (GBM) ultrastructurally. In contrast to Alport syndrome (A), there is no GBM thickening, lamellation, or granular inclusions. Morphologically, there is overlap between thin basement membrane nephropathy and A in female patients in whom there might be only thin GBM and no pathognomonic findings of A. Objective. To determine if the use of antibodies to collagen IV is helpful in making the distinction between thin basement membrane nephropathy and A in female patients with primarily thin GBMs. esign. We examined renal biopsies from 9 adult female patients with thin GBMs for the presence of 1, 3, 4, and 5 chains of type IV collagen by immunofluorescence. Results. In 2 patients with segmental GBM staining, no suggestion for A was found on physical examination or in their family history. In the remaining 7 patients with normal GBM staining, 4 had family members with end-stage renal disease of unknown etiology, raising the suspicion of X-linked or autosomal-recessive A. Three patients were presumed to have thin basement membrane nephropathy. Conclusion. egmental GBM staining for 3, 4, and 5 chains of type IV collagen raises the suspicion of A in the presence of adequate controls and other supporting evidence. Normal GBM staining for 3, 4, and 5 chains of type IV collagen, however, does not exclude A. (Arch Pathol Lab Med. 2001;125: ) The incidence of thin glomerular basement membrane nephropathy (TBMN; also called benign familial hematuria/nephritis) ranges between 5.2% and 9.2% in the general population 1 and may account for up to 30% of patients presenting with asymptomatic hematuria. 2 Thin glomerular basement membrane nephropathy (reviewed in references 3 and 4) is believed to have an autosomaldominant pattern of inheritance. A benign outcome is reported in the majority of patients, but glomerular obsolescence, proteinuria, and hypertension may occur. 5,6 The predominant ultrastructural anomaly in the kidney is a diffusely thin glomerular basement membrane (GBM). 2 In contrast to Alport syndrome (A), there are no widespread areas of thickening, lamellation, or granular inclusions within the GBM, 3,4 although focal splitting has been described. Alport syndrome can be X-linked, autosomal-recessive, or possibly autosomal-dominant (reviewed in references 3 and 7 10). There are a number of phenotypic variants of A, and the classic findings of hematuria, progressive renal failure, bilateral high-tone sensorineural hearing loss, and ocular changes are not found in all patients. 7 In the Accepted for publication ecember 7, From the epartment of Laboratory Medicine and Pathobiology, Toronto General Hospital, University Health Network, University of Toronto, Ontario. Reprints: Ginette Lajoie, M, FRCPC, epartment of Pathology, Eaton Wing 4-323, Toronto General Hospital, University Health Network, 200 Elizabeth t, Toronto, Ontario, M5G 2C4 Canada ( ginette.lajoie@uhn.on.ca). X-linked form, there is a defect in the gene coding for the 5 chain of type IV collagen ( 5[IV]) located on the long arm of chromosome X. Autosomal-recessive patients have mutation(s) in either COL4A3 or COL4A4 genes situated on chromosome 2. These 2 genes code for the 3 ( 3[IV]) and 4 ( 4[IV]) chains of type IV collagen, respectively. Ultrastructurally, the GBM presents alternating areas of thinning and thickening, splitting, lamellation, and granular inclusions in the lamina densa. These so-called pathognomonic findings are not always found in children and females, who may only have thin GBMs without other alterations. Therefore, it may be difficult to make a distinction between TBMN and A in these patients on a purely morphologic basis. Thus, the diagnosis of TBMN remains one of exclusion. Although it has been reported that glomeruli of patients with TBMN reveal no abnormalities when stained with antibodies against 3(IV), 4(IV), and 5(IV), 4 these observations have not been published. In the present study, we sought to examine the usefulness of antibodies against 1, 3, 4, and 5 chains of type IV collagen in establishing the diagnosis of TBMN in 9 adult female patients with diffusely thin GBMs. MATERIAL AN METHO Patient election and Clinical Information Nine female patients with diffusely thin GBM on renal biopsy were identified between 1994 and 1996 from the surgical pathology files of the University Health Network, Toronto, Ontario. Controls included 1 male patient with known X-linked A, 1 female patient with immunoglobulin A (IgA) nephropathy, and 2 Arch Pathol Lab Med Vol 125, May 2001 iagnosis of Thin Basement Membrane Nephropathy Lajoie 631

2 Patient No. Table 1. Clinical ummaries of 9 Female Patients Age at Biopsy, y Indication for Renal Biopsy Family History Follow-up 1 42 Microscopic hematuria for 3 y Two uncles have renal failure of unknown etiology; no deafness in family members 2 43 Potential donor for son; microscopic hematuria 3 36 Microscopic hematuria; proteinuria (0.4 g/24 h) 4 27 Microscopic hematuria; no proteinuria 5 37 Potential donor for brother; microscopic hematuria; no proteinuria 6 35 Proteinuria (1.0 g/24 h); microscopic hematuria 7 34 Potential donor for sister; hematuria 4 y; no proteinuria 8 23 Proteinuria (2.97 g/24 h) and hematuria 9 27 Increasing proteinuria (2.0 g/24 h at time of biopsy); hematuria since age 7 y on has renal failure due to reflux nephropathy; no other family history Mother, father, sister, and maternal aunt with hematuria No known family history of hematuria or renal disease Father has microscopic hematuria; brother has end-stage renal disease of unknown etiology No known family history of hematuria or renal disease ister with end-stage renal disease of unknown etiology; maternal aunt has microscopic hematuria Brother received kidney transplant at 40 y, primary disease unknown; 2 brothers and sisters have normal urinalysis No family history available; patient is adopted Normal renal function Normal renal function; history of goiter Increasing creatinine Increasing proteinuria during pregnancy (up to 7 g/24 h); 3 g/24 h postpartum; creatinine normal and stable Length of Follow-up 6mo 4mo 3y 3y Nil 4y 5y 18 mo 2.5 y female patients with minimal-change nephrotic syndrome. Clinical charts of index patients were reviewed for age, indication for renal biopsy, and family history. Clinical follow-up was obtained. Light, Immunofluorescence, and Electron Microscopy Renal biopsies were routinely processed for light, immunofluorescence, and electron microscopy. The glass slides were reviewed for all indexed cases and controls. All cases were examined by electron microscopy. The GBM was measured in all cases using the method of harmonic mean of orthogonal intercepts. 11 A measurement of less than 264 nm was required to make a diagnosis of TBMN. 2 Antibodies Murine monoclonal antibodies anti- 1(IV) (MAB1), anti- 3(IV) (MAB3), and anti- 5(IV) (MAB5-A7) were purchased from Wieslab AB (Lund, weden). The monoclonal antibody against 4(IV) (MAB85) was graciously provided by Clifford Kashtan (University of Minnesota, Minneapolis). Monoclonal antibodies against 1(IV) (MAB1), and 3(IV) (MAB3) were obtained from mice immunized with the noncollagenous C-terminal domain (NC1) of bovine GBM. 12 The monoclonal antibody against 5(IV) (MAB5-A7) was obtained from immunization of mice with the collagenase-resistant residue of human GBM. 13,14 The monoclonal antibody MAB85 (anti- 4[IV]) was obtained from mice immunized with human NC1. 15 Only the MAB85 was diluted 1:1 for immunohistochemistry. The other antibodies were not diluted. Immunofluorescence for 1(IV), 3(IV), 4(IV), and 5(IV) on Frozen Tissue The slides were stained for 1, 3, 4, and 5 chains of type IV collagen according to the method proposed by Yoshioka et al, 16 with slight variation. Briefly, slides were fixed in acetone for 5 minutes. After washing in phosphate-buffered saline, the slides were treated with a solution of 6 mol/l urea, 0.1 mol/l glycine atph3.5for1hourat4 C. Incubation with the primary antibodies for 45 minutes at room temperature followed. After a phosphate-buffered saline wash, fluorescein isothiocyanate conjugated goat anti-mouse antibody (ako Corporation, Carpinteria, Calif) was applied for 30 minutes. The slides were mounted in a permanent mounting solution (Gelvatol, Air Products and Chemicals, Allentown, Pa). The primary antibody was omitted in negative controls. REULT Clinical ummary The 9 female patients with diffusely thin GBMs were between 23 and 43 years old (average, 34 years) (Table 1). even patients presented with microscopic hematuria, with or without low-grade proteinuria; 3 of those patients were identified during assessment for living-related kidney donation. Two patients were being investigated because of proteinuria of 2.0 g/24 h or more. A family history of hematuria only was obtained in 1 patient. Two additional patients had a family history that included both hematuria and renal failure of unknown etiology. Another 2 patients had relatives with end-stage renal disease of unknown etiology, and 1 patient had a son diagnosed with reflux nephropathy. There was no known family history of hematuria or renal disease in 3 patients; 1 of these patients had been adopted. Clinical follow-up was obtained for 8 patients and varied from 4 months to 5 years (average, 26 months). Of the 8 patients for whom follow-up was available, 7 had stable renal function with persistent hematuria and 1 patient had progressive renal insufficiency. Light Microscopy No glomerular changes other than global and/or segmental glomerulosclerosis were observed (Table 2). Global glomerulosclerosis involved less than 10% of glomeruli in 3 patients, and 12% to 30% of glomeruli in 4 patients. Two 632 Arch Pathol Lab Med Vol 125, May 2001 iagnosis of Thin Basement Membrane Nephropathy Lajoie

3 Patient No. Light Microscopy Table 2. Light and Electron Microscopy* Electron Microscopy (Features Additional to iffuse Thinning of GBM) GBM Width Measurement, nm 9 12% global G and focal and segmental G; focal minimal IF 1 No G; no other LM changes Focal mild splitting % global G; focal mild IF None No G; focal mild IF None % global G; focal mild to moderate IF None % global G; no other LM changes Focal mild splitting % global G; focal minimal to mild IF None % global G; no IF None % global G and focal and segmental G; mild IF Focal mild splitting and rare intramembranous granu- 239 lar inclusions Focal mild splitting and rare intramembranous granular inclusions * GBM indicates glomerular basement membrane; G, glomerulosclerosis; LM, light microscopic; and IF, interstitial fibrosis. 179 Table 3. Immunofluorescence With Antibodies Against 1, 3, 4, and 5 Chains ( 1[IV], 3[IV], 4[IV], and 5[IV], respectively) of Type IV Collagen* Patient No. 1(IV) 3(IV) 4(IV) 5(IV) (weak) * indicates diffuse staining, as controls;, segmental staining. Figure 1. egmental sclerosis in a glomerulus from patient 8, who had normal staining of glomerular basement membrane with 3, 4, and 5 chains of type IV collagen (methylene blue, original magnification 100). of those 4 patients also had a few glomeruli with segmental scars (Figure 1). Interstitial fibrosis of moderate severity was seen in only 1 patient (patient 4) and was minimal or mild, and usually focal, in all other patients. Immunofluorescence Microscopy Immunofluorescence microscopy was negative for IgG, IgM, IgA, C3, C4, and and light chains in all cases. Immunofluorescence with antibodies against 1(IV), 3(IV), 4(IV), and 5(IV) (Table 3) showed diffuse and strong linear staining along GBM in 7 patients (Figure 2); this pattern was similar to that seen in positive controls (Figure 2). Patients 2 and 9 demonstrated segmental staining of GBM with 3, 4, and 5 chains of type IV collagen (Figure 2). Of those 2 patients, only patient 9 exhibited diffuse and strong linear staining for 1(IV) (Figure 3). Weak and uneven staining for 1(IV) was observed in the biopsy from patient 2. taining for 3, 4, and 5 chains of type IV collagen was lacking altogether in the male patient with X-linked A. iffuse and strong linear staining for 1(IV) was present in that patient (results not shown). Electron Microscopy In all cases, the mesangial matrix was either normal or mildly increased and did not contain electron-dense immune-type deposits (Table 2). Effacement of foot processes of glomerular visceral epithelial cells was noted over short segments. There was diffuse thinning of GBM that was confirmed by measuring orthogonal intercepts (Figure 4). The GBM measurement in indexed patients ranged from 161 to 239 nm (average, 200 nm), as opposed to controls, in which the GBM measured from 296 to 322 nm (average, 309 nm). The GBM of the male patient with A measured 324 nm and presented typical changes of A (namely, alternating areas of thin and thick GBM with splitting, lamellation, and granular inclusions). imilar changes were seen focally in only 1 indexed patient (Figure 5, A). Changes limited to splitting of short segments of GBM were seen in 3 cases (Figure 5, B). Notably, splitting of GBM over short segments was also seen in 2 cases of minimal-change nephrotic syndrome, and 1 case of IgA glomerulopathy. COMMENT A pathologist confronted with a renal biopsy in which diffuse GBM thinning is the predominant anomaly has to take numerous steps to try to establish an accurate diagnosis. In an attempt to rule out A, one must, if possible, obtain a complete clinical summary, including a detailed family history, and perform electron microscopy and immunofluorescence with antibodies against 3, 4, and 5 of type IV collagen. 17 Unfortunately, a detailed family history might be unobtainable 6 ; for example, patient 9 in our study was adopted. If and when available, the family history might be of limited use in some patients because of the possibility of a new mutation or because the etiology of renal failure in Arch Pathol Lab Med Vol 125, May 2001 iagnosis of Thin Basement Membrane Nephropathy Lajoie 633

4 Figure 2. Immunofluorescence staining for 3, 4, and 5 chains of type IV collagen. trong linear positive staining along the glomerular basement membrane is present in a control patient (A [ 3],B[ 4], and C [ 5], original magnification 300), and in patient 4, who was presumed to have thin basement membrane nephropathy ( [ 3], E [ 4], and F [ 5], original magnification 500). egmental staining is present in patient 9, suspected to have Alport syndrome (G [ 3], H [ 4], original magnification 312; I [ 5], original magnification 500). Figure 3. iffuse and strong linear staining for 1(IV) in patient 9 (original magnification 200). Figure 4. Renal biopsy from patient 2 showing diffusely thin glomerular basement membranes without other alterations (uranyl acetate and lead citrate, original magnification 3800). relatives is undetermined. Four patients in our study had relatives with renal failure of unknown origin. Furthermore, in a retrospective study like ours, and often in the practice of renal pathology, a complete clinical history cannot be obtained. Electron microscopy cannot discriminate between TBMN and A in patients in whom only a diffusely thin GBM with minimal alterations of the lamina densa is found. Any degree of GBM splitting or lamellation is compatible with A, even if the GBM changes are mild and present over short segments only. 17 Conversely, such changes are not pathognomonic for A and can be found in other renal diseases, 18 including TBMN. Among the 4 patients in our study with segmental splitting of GBM by electron microscopy (patients 1, 5, 8, and 9), only 1 patient (patient 9) had segmental GBM staining for 3, 4, and 5 chains of type IV collagen by immunofluorescence. This pattern of staining is likely specific for the diagnosis of A, but in the absence of other corroborative evidence (no stigmata of A and no family history), only genetic analysis would give a definitive diagnosis. 10 Another patient (patient 2), also with segmental GBM staining (but no GBM splitting), lacked clinical indices supporting a diagnosis of A, including a family history compatible with A. The absence of strong staining with 1 of type IV collagen in that patient indicates that the GBM antigens are not well preserved, and that the segmental staining observed with 3(IV), 4(IV), and 5(IV) is probably spurious. This case illustrates the need to perform adequate controls and to interpret the results with caution. Interestingly, the remaining 3 patients with segmental splitting of the GBM but with normal GBM staining with 3, 4, and 5 chains of type IV collagen all have male relatives with end-stage renal disease of unknown etiology. In these 3 patients, normal GBM staining does not exclude A or the carrier status of the Alport mutation. Favorable lyonization may result in a normal pattern of staining in females with X-linked A. Another possibility is that these patients are heterozygous for the mutation associated with autosomal-recessive A (see below). Furthermore, some 634 Arch Pathol Lab Med Vol 125, May 2001 iagnosis of Thin Basement Membrane Nephropathy Lajoie

5 Figure 5. A, Patient 9 with possible diagnosis of Alport syndrome. Electron microscopy reveals a segment of glomerular basement membrane (GBM) with lamellation (left) and another segment with marked thinning (right) (uranyl acetate and lead citrate, original magnification ). B, Patient 1. egment of GBM with focal splitting and normal staining of GBM with 3, 4, and 5 chains of type IV collagen. The patient has 2 uncles with end-stage renal disease of unknown etiology (uranyl acetate and lead citrate, original magnification 9500). patients with A have normal staining with antibodies against 3, 4, and 5 chains of type IV collagen. 19 Genetic studies could help sort out the likely pattern of inheritance of renal disease in these 3 families. Four patients (patients 3, 4, 6, and 7) demonstrated no ultrastructural GBM abnormalities other than diffuse thinning; all had a normal pattern of GBM staining by immunofluorescence. It is likely that these patients have TBMN. Of these 4 patients, 2 reported a family history of hematuria. Hematuria tends to be underreported unless family members are tested specifically. Renal function has remained normal and stable in these 4 patients during an average follow-up of 3.75 years, although patient 7 has a sister with end-stage renal disease of unknown etiology. In the family of this patient, the possibility of autosomalrecessive A has to be considered. Presumably, patient 7 would be heterozygous for the mutant gene (see below). In contrast, an extensive family history of hematuria would suggest TBMN with an autosomal dominant pattern of inheritance in patient 3. The putative protein defect in GBM of patients with TBMN remains unknown (reviewed in references 3 and 4). Based on linkage analysis studies, Lemmink et al 20 and others 21 have proposed that patients with TBMN might be heterozygous for the mutant gene that causes autosomalrecessive A. Their findings would tend to support the hypothesis that TBMN and A are entities belonging to the same spectrum of disease, although others have not been able to demonstrate a similar association. 22 Arguing against such a hypothesis is the fact that Nomura and his colleagues 23 reported no hematuria on repeated testing of heterozygous parents of a patient with autosomal-recessive A. Furthermore, in 3 animal models of autosomalrecessive A, heterozygous animals have a normal phenotype. 24 Thin GBM nephropathy is believed by some to have an autosomal-dominant pattern of inheritance. 3,4 The issue remains unresolved. In our study, no abnormality of 3(IV), 4(IV), or 5(IV) could be demonstrated by immunofluorescence in 7 patients. Possibly, this method lacks sensitivity for the detection of small quantitative changes. A more contentious issue is whether patients with TBMN should be considered for kidney donation. The long-term prognosis of patients thought to have TBMN is unknown. While a benign outcome is the rule, glomerular obsolescence, hypertension, and proteinuria have been reported. 5,6 It is important to note that the studies reporting progression toward renal insufficiency in TBMN might have unknowingly included unidentified A patients. It has been suggested that in patients with a presumed diagnosis of TBMN, only a detailed family history reporting several male members with long-standing hematuria and no progression toward renal failure in advanced age is the best evidence for a benign diagnosis. 7 Follow-up was too short for most of the patients included in our study to allow prognostication about renal outcome. Thin GBM in female patients presents a common diagnostic puzzle in the practice of renal pathology. There is much overlap between TBMN and A, especially in female patients. Our discussion emphasizes the need to correlate clinical history and pathologic findings in the interpretation of renal biopsies. The addition of immunofluorescence for 1, 3, 4, and 5 chains of type IV collagen might be useful in confirming a diagnosis of A in patients in whom there is a documented family history. 17 In other patients, abnormalities of GBM staining might raise the degree of suspicion for that diagnosis. In our study, patients presumed to have TBMN did not have detectable defects of 3, 4, and 5 chains of collagen type IV in the GBM. Normal GBM staining, however, does not exclude A or the carrier status of the Alport mutation; the family history of 4 of our patients with normal GBM staining was suggestive of A. tudies of families with suspected TBMN are needed to determine the pattern of inheritance of this disorder. There are possibly at least 2 groups of patients; the first group might be heterozygous carriers of the mutant gene for autosomal-recessive A, while in others TBMN might be inherited as an autosomal-dominant disease. This study has been funded in part by the University Health Network Pathology Associates and by the University Health Network epartment of Pathology, Toronto, Ontario. The author thanks Clifford Kashtan, M, for his generous gift of the MAB85 antibody and recognizes the excellent technical assistance of Richard Leung, Bc, ART (electron microscopy and photography) and Joanne Mariano, Bc, MLT (immunofluorescence). References 1. ische FE, Anderson VE, Keane J, Taube, Bewick M, Parsons V. Incidence of thin basement membrane nephropathy: morphometric investigation of a population sample. J Clin Pathol. 1990;43: Tiebosch ATMG, Frederik PM, van Breda Vriesman PJC, et al. Thin-basement-membrane nephropathy in adults with persistent hematuria. N Engl J Med. 1989;320: Arch Pathol Lab Med Vol 125, May 2001 iagnosis of Thin Basement Membrane Nephropathy Lajoie 635

6 3. Bodziak KA, Hammond W, Molitoris BA. Inherited diseases of the glomerular basement membrane. Am J Kidney is. 1994;23: Kashtan CE. Alport syndrome and thin glomerular basement membrane disease. J Am oc Nephrol. 1998;9: ische FE, Weston MJ, Parsons V. Abnormally thin glomerular basement membranes associated with hematuria, proteinuria or renal failure in adults. Am J Nephrol. 1985;5: Nieuwhof CMG, de Heer F, de Leeuw P, van Breda Vriesman PJC. Thin GBM nephropathy: premature glomerular obsolescence is associated with hypertension and late onset renal failure. Kidney Int. 1997;51: Gregory MC, Terreros A, Barker F, Fain PN, enison JC, Atkin CL. Alport syndrome: clinical phenotypes, incidence, and pathology. Contrib Nephrol. 1996;117: Kashtan CE, Michael AF. Alport syndrome: from bedside to genome to bedside. Am J Kidney is. 1993;22: Kashtan CE, Michael AF. Alport syndrome. Kidney Int. 1996;50: Pirson Y. Making the diagnosis of Alport s syndrome. Kidney Int. 1999;56: ische FE. Measurement of glomerular basement membrane thickness and its application to the diagnosis of thin-membrane nephropathy. Arch Pathol Lab Med. 1992;116: Johansson C, Butkowski R, Wieslander J. Characterization of monoclonal antibodies to the globular domain of collagen IV. Connect Tissue Res. 1991;25: Kleppel MM, Fan WW, Cheong HI, Kashtan CE, Michael AF. Immunochemical studies of the Alport antigen. Kidney Int. 1992;41: ing J, Kashtan CE, Fan WW, et al. A monoclonal antibody marker for Alport syndrome identifies the Alport antigen as the 5 chain of type IV collagen. Kidney Int. 1994;46: Kleppel MM, anti PA, Cameron J, Wieslander J, Michael AF. Human tissue distribution of novel basement membrane collagen. Am J Pathol. 1989;134: Yoshioka K, Michael AF, Velosa J, Fish AJ. etection of hidden nephritogenic antigen determinants in human renal and nonrenal basement membranes. Am J Pathol. 1985;121: Meleg-mith, Magliato, Cheles M, Garola RE, Kashtan CE. X-linked Alport syndrome in females. Hum Pathol. 1998;29: Hill G, Jenis EH, Goodloe. The nonspecificity of the ultrastructural alterations in hereditary nephritis. Lab Invest. 1974;31: Nakanishi K, Yoshikawa N, Iijima K, et al. Immunohistochemical study of 1 5 chains of type IV collagen in hereditary nephritis. Kidney Int. 1994;46: Lemmink HH, Nillesen WN, Mochizuki T, et al. Benign familial hematuria due to mutation of the type IV collagen 4 gene. J Clin Invest. 1996;98: Buzza M, Wilson, avige J. Linkage of thin basement membrane disease (TBM) to the loci for X-linked and autosomal recessive Alport syndrome [abstract]. J Am oc Nephrol. 1998;9:387A. 22. aito A, Yamazaki H, Nakagawa Y, Arakawa M. Molecular genetics of renal diseases. Intern Med. 1997;36: Nomura, Naito I, Fukushima T, et al. Molecular genetics and immunohistochemical study of autosomal recessive Alport s syndrome. J Am Kidney is. 1998;31:E Heikkilä P, Tryggvason K, Thorner P. Animal models of Alport syndrome: advancing the prospects for effective human gene therapy. Exp Nephrol. 2000;8: Arch Pathol Lab Med Vol 125, May 2001 iagnosis of Thin Basement Membrane Nephropathy Lajoie