—  SPECIALTY CONFERENCE  —

Nephropathology

Case 1 - Anderson-Fabry Disease

Robert B. Colvin
Massachusetts General Hospital
Boston, Massachusetts


Click on each slide thumbnail image for an enlarged view
Clinical History:
33 year old man presents with mild proteinuria (0.9 gm/d) and normal creatinine (0.8 gm/d). Since adolescence had paresthesias of hands and feet, most severe when febrile. He also has longstanding purple papules on flanks and groin area. No family history of renal disease. A renal biopsy (Biopsy A) was performed and he was entered into a clinical trial. At the end of the trial, 11 months later a second biopsy was performed (Biopsy B). Both biopsies were processed in epon for one micron sections and EM. What was the disease and treatment and how did the pathology change?

Biopsy
The first biopsy shows widespread granular, dense deposits in one micron sections that are most prominent in podocytes. On careful inspection (oil recommended) similar granular deposits can be seen of the glomerular and peritubular capillary endothelial cells (Figs. 1,2), arterial endothelial cells, the vascular smooth muscle, mesangial cells, interstitial macrophages/fibroblasts and in distal tubules. The electron micrographs reveal these are electron dense, amorphous and present in peritubular capillaries, arterial endothelium and vascular smooth muscle (Fig. 3). This pattern of podocyte deposition, accompanied by the other sites of lipid accumulation, is diagnostic of Fabry's disease, a glycosphingolipid storage disease.


Case 1 - Figure 1 - Biopsy A: Low power view showing lipid deposits in the podocytes, (toluidine blue).
Case 1 - Figure 2 - Biopsy A: The glomerulus shown here has lipid deposits in podocytes, endothelial cells, and mesangial cells (toluidine blue).
Case 1 - Figure 3 - Biopsy A: Peritubular capillaries show lipid within their endothelium (insert) (toluidine blue).


Case 1 - Figure 4 - Biopsy A: Electron microscopy of a peritubular capillary and arteriole show endothelial cell and smooth muscle lipid deposits.
Case 1 - Figure 5 - Biopsy B: Low power view showing lipid deposits in the podocytes, (toluidine blue).


Case 1 - Figure 6 - Biopsy B: On high power view the glomerulus shows little or no lipid in endothelium or mesangium. Podocytes still have lipid, (toluidine blue).
Case 1 - Figure 7 - Biopsy B: Electron micrograph of peritubular capillaries shows no lipid deposits remaining in the endothelium.

Diagnosis: Anderson-Fabry Disease

Clinical Features:
Anderson-Fabry's disease was described in the dermatologic literature in 1898 by Anderson in the UK and Fabry in Germany.1, 2  The characteristic purple papules were termed "angiokeratomas," although they were appreciated to be vascular ectasias, rather than tumors. A subsequent report by Fabry noted the development of renal failure in one of the original patients.3  Other well characterized features of Fabry's disease include acroparesthesias, hypohidrosis, cardiomyopathy, myocardial infarction, corneal/lenticular, opacities and thrombotic cerebrovascular disease. The major cause of death is renal disease, which characteristically develops in the 40s 4, 5 


Figure 1. ESRD in Fabry's Disease 4 

Pathogenesis:
A series of discoveries have lead to the molecular understanding of Fabry's disease as a deficiency of the lysosomal enzyme, a-galactosidase A, which metabolizes glycosphingolipids.6 

1898Angiokeratoma Anderson, Fabry
1930Renal failure Fabry
1947Vacuoles glom, vesselsPompen
1950Lipid in deposits Scriba
1963Globotriaosylceremide Sweeley, Kionsky
1965X linked recessive Opitz
1967a-galactosidase A def Brady
1980cDNA Bishop, Desnick
1989Gene Desnick, Bishop
2001Replacement therapy Eng/Desnick; Schiffman/Brady

The absence of the enzyme leads to the intracellular accumulation of globotriaosylceremide (GL-3) in endothelial cells, smooth muscle, macrophages and other cell types. Other substrates of the enzyme are also deposited such as B blood group substance in individuals with this blood group. The podocyte lipid is primarily galabiosylcerebroside.

The gene for a-galactosidase A is located on the X chromosome, hence the propensity for males to be affected. The mutation frequency is ~1:40,000 males. About 5% of cases are sporadic (not inherited). Renal outcome is related to the type of mutation and residual enzyme activity.7  Over 150 different mutations have been identified; some missense mutations have a milder disease.8  The "cardiac variant" has cardiac manifestation without renal disease failure and the "renal variant" lacks acroparesthesias, angiokeratoma, hypohidrosis, and corneal opacities but develops renal failure.9  As expected, renal disease is less common in women, however, it is notable that 12% of the Fabry''s patients on dialysis are women.5 

The accumulation of GL-3 and related glycolipids in various cells causes the clinical and pathological effects. Perhaps most important are the lipids in endothelial cells, believed to lead to microvascular obstruction and ischemia as well as an inflammatory and procoagulant response promoting microthrombosis .10, 11  The vascular smooth muscle lipid leads to microaneurysms and the angiokeratomas. Involvement of dorsal root ganglia lead to paresthesias and involvement of sympathetic ganglia and sweat glands to the hypohydrosis. In myocardial cells, the deposits cause defects in conduction and contraction. In distal tubules and collecting ducts the lipids lead to defective concentrating ability. The effect of the lipid in podocytes is uncertain. Mice that have the a-galactosidase A gene knocked out do not develop renal disease. These mice do not get the endothelial deposits, typical of humans, but do develop some podocyte lipid.12  This argues that the endothelial rather are more pathogenetic than the glomerular deposits.

Pathology:
The lipid, partially dissolved, is recognizable by light microscopy as clear vacuoles, particularly in podocytes and stains with luxol fast blue. In frozen sections the deposits are positive for PAS, oil red O, sudan black, bind the lectin Griffonia simplicifolia-I, are birefringent under polarized light and have a yellow autofluorescence in UV light.13  EM embedding preserves the lipid which appears as dense granules in one micron sections (Figure 1-2). By EM the deposits are round, dense, variegated and lamellated (myelin figures) in phagolysosomes, especially in the podocyte.14 (Figure 3). The concentric laminations have a periodicity of 3.5-5 nm .15  A "sunburst" pattern has also been noted in endothelial cells.16  Lipid droplets are common in podocytes in the setting of severe proteinuria, but these are not typically laminated. The differential of the laminated podocyte deposits is limited but does include silicon nephropathy.17 

In the kidney the most prominent site of accumulation is the podocyte, which looks characteristically "foamy" on H&E. The other conspicuous sites are distal tubules, collecting ducts and the vascular smooth muscle. Small lipid granules also are present diffusely in the vascular endothelium where they appear variously as individual single deposits to larger aggregates that bulge into the lumen. Additional features include focal segmental glomerular sclerosis, interstitial fibrosis, and tubular atrophy. Interstitial macrophages and fibroblasts also contain lipid.

Treatment:
Two recent randomized controlled clinical trials sponsored by Genzyme and Transkaryotics have tested enzyme replacement therapy .18, 19  In these studies pre- and post-treatment renal biopsies were used to assess efficacy. In both, decreased plasma lipid and tissue deposition was demonstrated.

The Genzyme study used recombinant human a-galactosidase A secreted from CHO cells, (rh- aGAL, Fabrazyme), which contains mannose-6-phosphate (the intracellular signal for trafficking to lysosomes).18  The primary endpoint was loss of endothelial lipid from peritubular capillaries after 20 weeks of treatment, as judged by 3 independent renal pathology consultants (Helmut Rennke, Steven Dikman, and myself) in one micron sections under oil, scoring and counting each PTC without knowledge of the treatment group or timing of the biopsy. In addition, endomyocardial and skin biopsies were scored similarly by other pathologists. The patients (n=58) had no or minimal renal impairment (mean Cr 0.8mg/dl) and were randomly assigned to receive placebo or 1 mg/kg of rh-aGAL i.v. q14 days. The results were striking, in the kidney, heart and skin capillaries, dramatic reductions of endothelial lipid was demonstrated compared with the placebo controls (p<.001). In the kidney the primary endpoint was a score of 0 in >50% of the PTC and <5% with a score >0, achieved by 20/29 of the treated and 0/29 of the control patients. The drug had no adverse effects that stopped treatment, although 88% of the patients developed antibodies to rh- aGAL. Subsequently all patients received rh-aGAL for another 6 months and repeat biopsies were done.20 

Detailed analysis of the three sequential biopsies showed that lipid was cleared completely from all endothelial cells (PTC, glomerular, arterial), mesangial cells, interstitial cells and to a lesser extent tubules and vascular smooth muscle.20  Podocytes were the most resistant to clearance. The reason for the difference is not certain. Among the possibilities are that 1) the podocytes accumulate a different lipid (galabiosylcerebroside), but rh-aGAL digests this in vitro; 2) podocytes have a slower turnover than endothelial cells; or 3) the enzyme, with a MW of ~100 kd, has limited access to podocytes. Whether podocyte lipid is important or not to the pathogenesis of renal disease is uncertain. The mouse knockout has no renal disease but does have mild podocyte lipid. The mouse KO lacks endothelial deposition, however, adding credence to the relevance of that site to the pathogenesis.

Even though the patients typically make anti-rh-aGAL antibodies, no evidence of immune complex renal disease was found (except for 2 cases of pre-existing IgA deposition).20 

The present case was in the rh- aGAL study and assigned to the treatment group. Biopsy (A) is pretreatment and biopsy (B) is after 11 months of rh-aGAL. The followup biopsy shows marked loss of endothelial, mesangial and interstitial lipid accumulation (Figures 4-6). It is hard to be certain of any change in tubules, smooth muscle or podocytes.

Next Steps:
The goal of treatment is to prevent or reverse renal disease as well as the other manifestations. Studies are in progress that will determine the effect of enzyme replacement on renal function. At this time (December 2002) the FDA has not approved either a-Galactosidase A preparation, although both have been approved in Europe.

Acknowledgement:
The case material was kindly provided by Dr. Beth Thurberg (Genzyme) for this teaching exercise. The history that was given, except for the age and sex, was a "fabrycation" to protect patient confidentiality.

References

  1. Anderson WA, A case of angiokeratoma. Br J Dermatol, 10: 113, 1898.
  2. Fabry J, Ein beitrag zur kennitnis der purpura haemorrhagica nodularis. Arch Dermatol Syph, 43: 187, 1898.
  3. Fabry J, Dermatol Wochenschr, 90: 339, 1930.
  4. Branton MH, et al., Natural history of Fabry renal disease: influence of alpha- galactosidase A activity and genetic mutations on clinical course. Medicine (Baltimore), 81: 122-38., 2002.
  5. Thadhani R, et al., Patients with Fabry disease on dialysis in the United States. Kidney Int, 61: 249-55., 2002.
  6. Desnick RJ and Wasserstein MP, Fabry disease: clinical features and recent advances in enzyme replacement therapy. Adv Nephrol Necker Hosp, 31: 317-39, 2001.
  7. Schiffmann R, Natural history of Fabry disease in males: preliminary observations. J Inherit Metab Dis, 24: 15-7; discussion 11-2., 2001.
  8. Ashton-Prolla P, et al., Fabry disease: twenty-two novel mutations in the alpha-galactosidase A gene and genotype/phenotype correlations in severely and mildly affected hemizygotes and heterozygotes. J Investig Med, 48: 227-35, 2000.
  9. Desnick RJ, Banikazemi M and Wasserstein M, Enzyme replacement therapy for Fabry disease, an inherited nephropathy. Clin Nephrol, 57: 1-8., 2002.
  10. DeGraba T, et al., Profile of endothelial and leukocyte activation in Fabry patients. Ann Neurol, 47: 229-33., 2000.
  11. Utsumi K, et al., High incidence of thrombosis in Fabry's disease. Intern Med, 36: 327-9., 1997.
  12. Takahashi H, et al., Long-term systemic therapy of Fabry disease in a knockout mouse by adeno-associated virus-mediated muscle-directed gene transfer. Proc Natl Acad Sci U S A, 99: 13777-82., 2002.
  13. Alroy J, Sabnis S and Kopp JB, Renal pathology in Fabry disease. J Am Soc Nephrol, 13 Suppl 2: S134-8., 2002.
  14. Schatzki PF, Kipreos B and Payne J, Fabry's disease. Primary diagnosis by electron microscopy. Am J Surg Pathol, 3: 211-9., 1979.
  15. Sessa A, et al., Renal pathological changes in Fabry disease. J Inherit Metab Dis, 24: 66-70; discussion 65., 2001.
  16. Elleder M, et al., An atypical ultrastructural pattern in Fabry's disease: a study on its nature and incidence in 7 cases. Ultrastruct Pathol, 14: 467-74., 1990.
  17. Banks DE, et al., Silicon nephropathy mimicking Fabry's disease. Am J Nephrol, 3: 279-84, 1983.
  18. Eng CM, et al., Safety and efficacy of recombinant human alpha-galactosidase A-- replacement therapy in Fabry's disease. N Engl J Med, 345: 9-16., 2001.
  19. Schiffmann R, et al., Enzyme replacement therapy in fabry disease: a randomized controlled trial. JAMA, 285: 2743-9., 2001.
  20. Thurberg BL, et al., Globotriaosylceramide accumulation in the Fabry kidney is cleared from multiple cell types after enzyme replacement therapy. Kidney Int, 62: 1933-1946., 2002.