Detection of Pi Z phenotype individuals by alpha-1-antitrypsin (AAT) immunohistochemistry in paraffin-embedded liver tissue specimens

Abstract

Immunostaining for alpha-1-antitrypsin (AAT) was applied on paraffin-embedded liver specimens from 38 Pi Z phenotype (27 Pi MZ; 9 Pi ZZ; 2 Pi SZ) individuals (33 adults, 5 newborns). Histological diagnoses included normal liver, various forms of chronic hepatitis, cirrhosis. A positive hepatocytic staining for AAT was revealed in all cases, whereas on PAS-D staining 5 cases were negative. Immunostaining revealed different patterns: type I: fine solid granules or coarse laminated globules filling the whole liver cell cytoplasm; type II: positive granules marginated towards the cell's periphery; type III: positive granules restricted to focal areas in the cytoplasm. Type I and III patterns were observed in all specimens. Association of type I and III with type II was seen in all Pi MZ and Pi SZ cases, but only in 3 of the 9 Pi ZZ specimens. To check the specificity of these staining patterns for the Z allele of AAT in adults and newborns, 180 further cases were investigated blindly by immunohistochemistry. Recognition of hepatocytic staining patterns type I, II and III allowed to reliably diagnose the Pi Z phenotype in 6 cases, whose heterozygous state was confirmed by independent serum phenotyping (5 Pi MZ, 1 Pi SZ). These results allow to conclude that: (1) immunohistochemistry is a reliable technique (superior to PAS-D staining) to identify Pi Z individuals (homo- and heterozygotes) on paraffin-embedded liver sections; (2) heterozygotes Pi MZ and Pi SZ cannot be differentiated without serum phenotyping; (3) immunohistochemistry is helpful in distinguishing between Z homozygotes from heterozygotes, but determination of AAT serum concentration is required to allow confident distinction without phenotyping. AAT immunohistochemistry thus allows to diagnose the Z allele of AAT when phenotyping procedures are not available or in retrospective studies on biopsy and autopsy material.