Histologic, ultrastructural, chemical, and statistical methods were used to study liver biopsy and autopsy specimens from 43 patients who had primary sclerosing cholangitis (PSC), with or without chronic ulcerative colitis (CUC), and from 19 patients who had CUC without PSC. In all study groups, essentially the same abnormalities were found in the hepatic parenchyma outside the major bile ducts, although nondiagnostic tissue samples were observed also. Specimens from patients with extrahepatic PSC were indistinguishable from those patients with combined extra- and intrahepatic PSC. Common findings included periductal fibrosis and inflammation, portal edema and fibrosis, focal proliferation of bile ducts and ductules, focal bile duct obliteration and loss of bile ducts, copper deposition, and cholestasis. Proliferation of bile ducts in some portal tracts and obliteration or absence of bile duct in others were the most characteristic changes. In most specimens, inflammatory changes appeared mild, yet biliary cirrhosis had developed in 34% of the patients. Specimens from patients with PSC, with or without CUC, more often contained bile and strikingly increased stainable copper (Grades 2 and 3) than did specimens from patients with CUC without PSC. Hepatic copper contents, measured by atomic absorption spectrophotometry, also were higher in specimens from patients with PSC. Study of PCS specimens by transmission electron microscopy and by energy-dispersive X-ray microanalysis revealed that most copper was sequestered in lipolysosomes. The recognition of strikingly similar morphologic features in many liver specimens from patients with either PSC or CUC or both suggests that the causes of these conditions are closely related.

In each heart taken from autopsies of 14 men with idiopathic hemochromatosis, the conduction system, atria and 10 sites in the ventricles were histologically graded for stainable iron. Stainable iron was exclusively sarcoplasmic; none was observed in the interstitium. The histologic grade for the same anatomic site varied among hearts and among different anatomic sites in the same heart. Ten hearts had stainable iron in all ventricular sites; one of the three hearts from patients who had undergone therapeutic phlebotomy had no iron at any site. Seven hearts had iron in the atria but at a lesser grade than that found in the ventricles; six hearts had mild focal iron deposition in the atrioventricular conduction system. None of the 14 hearts had stainable iron in the sinus node. Elemental iron was quantitated by atomic absorption spectroscopy in ventricular specimens contiguous to those studied histologically and also in age-matched control hearts. Elemental iron content was markedly increased in hearts with idiopathic hemochromatosis compared with control hearts (p less than 0.01). The quantity of elemental iron varied greatly, similar to stainable iron, but was highest subepicardially. Among the hearts from the 11 patients without prior phlebotomy, three had no stainable iron in the right ventricular septal subendocardium, suggesting that sampling error may be a problem in the evaluation of hemochromatosis by endomyocardial biopsy. The sarcoplasmic location of the iron indicates that cardiac involvement in idiopathic hemochromatosis represents a storage disease and not an infiltrative process; this finding is consistent with the normal ventricular wall thicknesses observed.

Published figures for trace element concentrations in body fluids and tissues of apparently healthy subjects are widely divergent. For a considerable time, the apparent disparities were readily ascribed to biological sources of variation such as age, sex, dietary habits, physiological conditions, environmental exposure, geographical circumstances, or similar influences. Growing evidence, however, suggests that this interpretation may be seriously questioned in numerous instances. First, values obtained in reference materials leave no doubt that some previous studies must have been subject to gross analytical inaccuracies. Second, it has now been thoroughly documented that inadequate sample collection and manipulation may drastically distort the intrinsic trace element content of biological matrices. This review scrutinizes data reported by a number of investigators. In an effort to settle the currently flourishing confusion, critically selected reference values are set forth for trace element levels in human blood plasma or serum, packed blood cells, urine, lung, liver, kidney, and skeletal muscle tissue.

Eleven Bedlington terriers were found to have a mean hepatic copper concentration of 6,321 micrograms/g dry wt (normal, 200 micrograms/g dry wt) and renal copper concentration that was three or four times normal. Brain copper levels were normal in younger dogs, were elevated in two older dogs, and were 100 times normal in one dog that died of the disease. Increased concentrations of copper in the liver, kidney, and brain also characterize Wilson's disease. Erythrocyte survival was normal in three affected dogs, but serum glutamic-pyruvic transaminase levels were usually elevated. Unlike the hypoceruloplasminemia of patients with Wilson's disease, plasma ceruloplasmin activity was not only normal but was also slightly elevated in the terriers. Despite their normal or excessive ceruloplasmin, the Bedlington terriers could convert ionic 64Cu to radioceruloplasmin but did so only very slowly. These dogs accumulated significantly more 64Cu in their livers than normal, much like patients with Wilson's disease do before symptoms develop.