Light and electron microscopic features of the liver in mucopolysaccharidosis

Light and electron microscopic features of the liver in mucopolysaccharidosis

light and Electron Microscopic Features of the Liver in Mucopolysaccharidosis JEFFREY M. RESNICK, MD, CHESTER B. WHITLEY, PHD, MD, ARNOLD S. LEONARD, ...

9MB Sizes 0 Downloads 73 Views

light and Electron Microscopic Features of the Liver in Mucopolysaccharidosis JEFFREY M. RESNICK, MD, CHESTER B. WHITLEY, PHD, MD, ARNOLD S. LEONARD, MD, PHD, WILLIAM KRIVIT, MD, PHD, AND DALE Co SNOVER, MD Themucopolysaccharidosi (ME’S) diseaseslead to the accumulation of glycosaminoglycan in manytissues.In thisstudy 19 MI’S I, one MPS II, five MPS III, and two MPS VI patientsunderwentliver biopsy for ligbt and electron microscopicexamination.Electronmicroscopywas performed for all 27 specimens. Twemy-sixspecimens were studied by light microscopy, and the slides were stained with colloidal iron and alcian blue in 26 and six biopsy specimens, respectively.By hemato+ineosin stain 20 of 26 cases showed bepatocellulardilatation with rarefactionof the cytoplasm; the Kupffer cells were unremarkable. Twenty-fourand 25 of the 26 biopsy specimensshowed substantial colloidal iron stainingof hepatocytesand Kupffer cella, respectively. The six biopsy specimens prepared with al&n blue stain showed no reactivityof any cell type. Electron microscopy revealed characteristicmembrane-boundinclusionswithinthe hepatocytesand Kupffer cells of all 27 biopsy specimens.Of 19 cases in whichIto cells were identified, 18 included cells containing similar inclusions. Twentyof 27 biopsy specimens also demonstratedthe hepatocelhtbu accumulationof lipid droplets. Although there were no absolute distinguishingfeaturesamong the variousMPS diseases,the two MPS VI cases showed glycosaminoglycaninclusionsthat were fewer in number, smaller, and containedmore abundantlipofuscin than those a+ sociated with tbe other hfPS types. HUM PATHOL 25:276-286. Copyrlgbt 0 1994 by W.B. Saunders Company

The mucopolysaccharidosis (MPS) diseases represent a group of recessively inherited lysosomal storage diseases characterized by specific enzyme deficiencies, tissue accumulation of glycosaminoglycans (GAGS), and somatic deformities, neurologic deterioration, and mental retardation of varying severity.’ Hurler syndrome (MPS I), the prototypic MPS disease, typically leads to progressive debility and death before adolescence. Currently, allogeneic bone marrow transplantation (BMT) represents the only interventional modality available.*,’ Numerous case reports and small series have described the morphologic features of various tissues ob mined by autopsy or biopsy of patients with MPS.427 Because many patients manifest hepatomegaly and mental retardation, most attention has been focused on the pathologic changes of the liver and brain. Nevertheless, the largest reported series addressing hepatic morFrom the Departments of Laboratory Medicine and Pathology, Pediatrics, and Surgery, University of Minnesota Hospitals and Clinics, and the Institute of Human Genetics, University of Minnesota Medical School, Minneapolis, MN. Accepted for publication October 4, 1993. Z+y wmd.r: mucopolysaccharidosis, glycosaminoglycan, liver, colloidal iron stain, electron microscopy. Address correspondence and reprint requests to Dale C. Snover, MD, Department of Laboratory Medicine and Pathology, University of Minnesota Hospitals and Clinics, Box 76, UMHC, 420 Delaware St SE, Minneapolis, MN 55455. Copyright 0 1994 by W.B. Saunders Company 0046-8177/94/2503-0011$5.00/0

276

phology in MPS consisted of biopsies from 11 children with MPS I.** Since the biochemical classification of the MPS diseases was not established until the 196Os, many cases reported as Hurler syndrome may represent other MPS diseases. Furthermore, although the literature contains many accounts of the light microscopic (LM) examination of the liver in MPS, relatively few electron microscopic (EM) descriptions appear.8,‘1-12,15-16,z1~23 With an accepted and reliable classification system of the MPS diseases, it is appropriate to assemble a more complete series of liver biopsies to assess the pathologic changes in MPS. Because of the active BMT program at the University of Minnesota Hospitals, a relatively large number of MPS patients have been referred to our institution. Since the liver is well documented to accumulate GAGS in MPS, is readily accessible for biopsy, and comprises several different cell types, we chose to examine liver biopsy specimens by LM and EM. Our aim in this study was to define the spectrum of the morphologic changes occurring in MPS, to determine the features characteristic of specific MPS diseases, to identify the methodologies by which the pathology can best be demonstrated, and to consider the possible role for liver biopsy in the diagnosis of MPS. MATERIALS

AND METHODS

Liver biopsy specimens from 26 needle biopsies and one incisional biopsy were obtained from 27 patients with MPS (19 with MPS I [Hurler syndrome], one with MPS II [Hurler syn-

drome] , three with MPS III [Sanfilippo syndrome] type A, one with MPS III type B, one with MPS III type C, and two with MPS VI [Maroteaux-Lamy syndrome]). Twenty-six of the 27 patients underwent liver biopsy before BMT. The biopsy specimen from one MPS VI patient (unique patient no. [UPN] 217) was obtained 1 month after BMT (Table 1). These patients had been diagnosed definitively by clinical assessment and by laboratory measurement of quantitative urine total GAG excretion and leukocyte enzyme activity.28329 One patient (UPN 659; Table 1) was diagnosed unusually early with MPS IIIB because of the incidental finding of hepatic GAG accumulation during an evaluation of failure to thrive. All 27 patients’ pretransplant serum aspartate aminotransferase, alanine aminotransferase, 5’ nucleotidase, and bilirubin concentrations were within normal limits. Portions of the 26 biopsy specimens were processed for both LM and EM examination; one specimen had been sub mitted for ultrastructural assessment only (UPN 729). The tissue submitted for LM was fixed in 10% neutral phosphatebuffered formalin, embedded in paraffin, sectioned at a thickness of 5 pm, and stained with the hematoxylineosin, trichrome, reticulin, and colloidal iron stains.50 Prussian blue

THE LIVER IN MUCOPOLYSACCHARIDOSIS TABLE

1.

Summary of Patients and Their Ages Time of Liver Biopsy

On ultrastructural examination, the GAG inclusions of hepatocytes, Kupffer cells, and Ito cells were examined by both qualitative and semiquantitative parameters. For each biopsy the spectrum of the inclusions’ features and the predominant inclusion type were noted. For each patient the mean number of hepatocellular inclusions per photograph was calculated and averaged to a standard magnification of X7,490. The range of inclusion diameters also was recorded. The Kupffer and Ito cell GAG inclusions were examined similarly. For each cell type the average number of inclusions per cell and inclusion diameter range were determined.

at the

Unique Patient No.*

Disease

Age at Biopsy (yr)

819 1101 618 793 677 1040 647 1089 628 1310 1223 977 330 699 729 1104 453 463 926 759 950 1306 503 659 1149 1130 217

MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS I MPS II MPS III4 MPS III4 MPS III4 MPS IIIB MPS IIIC MPS VI MPS VI

0.9 1.0 1.0 1.1 1.3 1.3 1.6 1.7 1.7 1.8 1.9 1.9 2.1 2.2 2.4 2.6 2.8 3.1 3.4 4.4 2.9 5.6 5.8 1.9 7.1 1.6 12.9

(Resnick et al)

RESULTS The study included a patient population of 27, ranging in age from 0.9 to 12.9 years (median age, 1.9 years) at the time of liver biopsy (Table 1). Light

* From the Universityof Minnesota bone marrowtransplantation database.

stain served as negative control to rule out the presence of excess endogenous iron. We have found that liver biopsy specimens from patients with MPS who have achieved metabolic correction by BMT do not stain with colloidal iron.s Six of the biopsy specimens also were stained with alcian blue at pH 2.5 (UPNs 330,628,659,759,1306, and 1310).5’ Our laboratory’s general experience has been that colloidal iron is more sensitive in detecting excess GAGS than is the alcian blue stain. The tissue intended for EM was fixed in 2.5% glutaraldehyde solution (in Millonig’s buffer), postfixed in 1.5% osmium tetroxide solution, dehydrated, and embedded in Polybed resin. Ultrathin sections were stained with lead citrate and uranyl acetate. These sections were searched for hepatocytes, Kupffer cells, and Ito cells, and fields to be photographed for further examination (ranging in magnification from X3,920 to ~25,660) were selected randomly to assure representative sampling of the tissue. All biopsy specimens were examined independently by two pathologists (J.M.R. and D.C.S.) and randomly, without regard to specific MPS disease or patient age. The LM slides were examined for hepatic architectural alterations and the cytologic features of hepatocytes, sinusoidal cells (predominantly Kupffer cells), bile ductal epithelium, and vascular endothelium. In the assessment of these cells for reactivity with colloidal iron and alcian blue, only the presence of intracellular discrete particulate or deep homogenous staining was considered to be positive. For each cell type the extent of staining was graded semiquantitatively on scales of distribution and individual cell intensity. The percentage of colloidal iron-positive cells was estimated and categorized within a distribution grade of 0 (0 to 1%)) 1 (> 1% to 33%)) 2 (>33% to 66%), or 3 (>66% to 100%). The intensity of staining in a typical cell was assessed and graded as A (negligible) , B (slight), C (moderate), or D (strong).

FIGURE 1. Hepatocytes (top) and interlobular bile duct (bottom) in two cases of MPS. The hepatocytes (top) are distended and show cytoplasmic rarefaction. As a result of the distension many sinusoids are obliterated and the cell membranes of adjacent hepatocytes apparently are fused. The interlobular bile duct (bottom) shows cytoplasmic vacuolization with basal displacement of the nuclei. (Top: UPN 503: hematoxylin-eosin stain; original magnification x100. Bottom: UPN 330: hematoxYlin-eosin stain; original magnification x300.)

277

HUMAN

PATHOLOGY

Volume 25, No. 3 (March 1994)

1% 16 s

14

4

12

p

10

q

No Hepatocellular

q

Mild

w

Marked

Dilatation

Hepatocellular

Dilatation

Hepatocellular

Dilatation

i: *

4

FIQURE 2. As demonstrated by thls bar graph, 20 of 26 liver biopsy specimens showed at least sllght hepatocellular dllatatlon.

2

+

0 MPS I

MPS II

MPS III

MPS VI

3B), two were grade 2 (one grade 2C and one grade 2B), and one was grade OB (slight staining of a rare cell). The MPS II patient (UPN 759) showed grade 3C staining of the hepatocytes. The three MPS IIIA cases were grade 3C. The MPS IIIB and IIIC biopsy specimens were grades 2B and 2C, respectively. The two MPS VI patients showed grades 2B and OB staining patterns, respectively. Although 19 of the 26 cases showed grade 3 distribution of hepatocellular colloidal iron positivity (>66% to 100% of cells), the intensity of individual cell reactivity varied among these cases, ranging from grades B to D (slight to strong). Figure 4 illustrates the range of hepatocellular staining intensity among the MPS cases. Nevertheless, those biopsy specimens showing the higher percentages of positive cells tended to show more intense individual cell staining. Specimens from only one patient with MPS I (UPN 1040) and one with MPS VI (UPN 217) demonstrated no substantial hepatocellular colloidal iron positivity. The MPS I patient was not clinically distinct. However, the MPS VI patient was biopsied 1 month after BMT. The patterns of Kupffer cell reactivity for colloidal iron were categorized with the same grading system ap plied to the hepatocytes. Of the 18 MPS I cases, eight were grade 3D, five were grade 2D, four were grade 1 (two grade lD, one grade lC, and one grade lB), and one was grade OB. The one MPS II case showed grade 2D staining. Among the three MPS IIIA cases, two were grade 3D and one was grade 2D. The two patients with

microscopy examination revealed no architectural changes or inflammatory infiltrates. By hematoxylineosin stain, 20 of 26 biopsy specimens showed apparent distension of hepatocytes with cytoplasmic attenuation. The bloated hepatocytes were juxtaposed with apparent fusion of adjacent cell membranes and obliteration of adjacent sinusoids, resulting in a mosaic-like pattern (Fig 1, top). This effect was generally more prominent periportally. Although two cases (UPNs 659 and 217) showed steatosis, characterized by cytoplasmic vacuolization with nuclear displacement, the hepatocytes of the remaining biopsies revealed only cytoplasmic rarefaction (Fig 1, top). Figure 2 shows the proportion of cases demonstrating these cytologic features. By hematoxylineosin stain the sinusoidal cells (predominantly Kupffer cells) were unremarkable, displaying no vacuolization. Fifteen of 26 cases showed cytoplasmic clearing of the interlobular bile ductal epithelial cells (Figs 1, bottom and 3). Although no cases showed extensive fibrosis, a small central scar was present in the biopsy specimen from one patient with MPS VI (UPN 217). Only two biopsy specimens, which were obtained from patients with MPS I (UPNs 1223 and 1310), showed no pathologic changes detectable by hematoxylineosin stain. These two cases were not distinctive clinically. The biopsy specimens showed variable patterns of hepatocellular staining with colloidal iron. Of 18 biopsy specimens obtained from patients with MPS I, 15 were grade 3 (two grade 3D, seven grade 3C, and six grade 10

d g z p B

16

q

No Cytoplasmic

14

n

Cytoplasmic

Clearing

Clearing

12

FIQURE 3. As demonstrated by this bar graph, 15 of 26 liver biopsy specimens showed cytoplasmlc clearing of the interlobular bile ductal epithelium.

10

%

2

f 2

6 4 2

MPS I

MPS II

MPS Ill Dhase

278

MPS VI

FIGURE 4. Liver biopsy specimens from two patients with MPS I. (Top) Strong hepatocellular staining with colloidal-iron and relatively few colloidal-iron-positive Kupffer cells. (A Kupffer cell is labeled with an arrow.) (Bottom) Widespread Kupffer cell and negligible hepatocellular colloidal-iron reactivity. Note the differences between the two biopsy specimens in the staining intensity of hepatocytes and the proportion of colloidal-iron-positive Kupffer cells. (Top: UPN 1089: colloidal-iron staln; original magnification x66. Bottom: UPN 1101; colloidal Iron stain; original magnification x66.)

FIQURE 5. This large hepatlc artery is lined by colloidal-Iron-positive endothelium. The adjacent septal bile duct normally contains mucin and stains Intensely with colloidal iron. (UPN 1089; colloidal-iron stain; original magnification x66.)

279

HUMAN

PATHOLOGY

Volume 25, No. 3 (March 1994)

MPS IIIB and IIIC showed grade ID and IB staining, respectively. The two MPS VI patients were grades 3D and lD, respectively. Although 22 of the 26 patients showed grade D individual Kupffer cell staining intensity, the percentage of colloidal-iron-positive Kupffer cells was much more variable among the cases (Fig 4). As demonstrated in Fig 4, Kupffer cells tended to stain intensely and homogeneously or not at all. Biopsy specimens varied in the proportion of colloidal-iron-positive Kupffer cells. Only one MPS I case (UPN 1310) showed a negligible number of slightly staining Kupffer cells (grade OB); interestingly, the hepatocytes in this biopsy specimen showed grade 3B staining, indicating faint staining in more than 66% of the cells. Three cases (two with MPS I and one with MPS IIIC) showed slight individual Kupffer cell reactivity. Since these three biopsy specimens with slight staining intensity were of grades 1, 0, and 1 in the proportion of colloidal-ironpositive Kupffer cells, the specimens with the smallest proportion of colloidal-iron-positive cells showed the faintest individual cell reactivity. All six biopsy specimens containing septal bile ducts (including four MPS I, one MPS II, and one MPS IIIA cases) showed staining of their epithelium with colloidal iron (Fig 5). The interlobular bile ductal epithelium showed cytoplasmic clearing without colloidal-iron reactivity in all 15 cases. One biopsy (UPN 1089) showed endothelial cell staining for colloidal iron. This endothelium belonged to a hepatic artery in a large portal tract. The staining was finely particulate and involved the complete ring of lining endothelial cells (Fig 5). The vascular media and adventitia showed no reactivity. Of six biopsy specimens for which both colloidal iron and alcian blue histochemical techniques were performed (including three MPS I, one MPS II, one MPS IIIA, and one MPS IIIB cases), none showed detectable hepatocyte or Kupffer cell staining by alcian blue. However, all six cases showed substantial colloidal iron reactivity for both cell types. The EM examinations revealed hepatocytes containing characteristic single membrane-bound vacuoles. These hepatocellular lysosomal inclusions from all 27 biopsies comprised flocculent-reticular material of variable electron density (Fig 6). Many such inclusions contained well-circumscribed, electrondense circular bodies (Fig 6). Inclusions containing small lipofuscin particles and myelin figures also were identified (Fig 6). Although the individual inclusions were circular, several appeared to have been in the midst of coalescing. The two MPS VI cases (UPNs 1130 and 217) showed abundant lipofuscin particles within the inclusions (Fig 7). Biopsy specimens from 13 MPS I, five MPS III, and the two MPS VI cases showed considerable amounts of neutral lipid in the cytoplasm of hepatocytes. No mitochondrial pleomorphism was appreciated. The hepatocellular GAG inclusions also were evaluated quantitatively. Table 2 summarizes the number and size of inclusions for the MPS diseases. The hepatocytes of the two MPS VI patients contained far fewer GAG inclusions than those of the other MPS diseases

FIGURE 6. Hepatocyte cytoplasm from a patient with MPS IllA demonstrating abundant inclusions containing flocculent material and prominent electron-dense circular bodies (arrow). Myelin figures (arrowhead) also appear inside several inclusions (UPN 503; original magnification xl 1,440.)

(Fig 8). In addition, the hepatocytes of one MPS VI patient (UPN 217) contained a disproportionate number of lipid droplets, outnumbering the GAG inclusions. Corresponding to this biopsy specimen’s ultrastructural appearance was the light microscopic evidence of fatty change. Despite the comparable ranges of hepatocellular inclusion sizes among the MPSs (Table 2), the inclusions in MPS VI were substantially smaller than the typical inclusions associated with the other MPS diseases. Among the 27 cases, no correlation of inclusion number or size with age could be demonstrated (Fig 9). In the comparison of ultrastructure and LM with colloidal iron stain only two discrepant cases (with MPS I [UPN 10401 and VI [UPN 217]), showing minimal staining by LM and substantial numbers of GAG inclusions by EM, were noted. Interestingly, although 18 cases showed abundant neutral lipid ultrastructurally, only two biopsy specimens (one MPS IIIB case [UPN 6591 and one MPS VI case [UPN 2171) demonstrated steatosis at the LM level. From two to 26 (median, eight) Kupffer cells for each of the 27 biopsy specimens were identified for ultrastructural study. The Kupffer cells of all 27 biopsy specimens contained membrane-bound inclusions consisting of flocculent-reticular material. Intrainclusion electron-dense bodies frequently were present in the Kupffer cell inclusions, as was the case with hepatocytes. The lipofuscin particles and myelin figures appearing within the hepatocellular inclusions were found only rarely among the Kupffer cell inclusions. The Kupffer cell GAG inclusions demonstrated no features by which to distinguish the various MPS diseases from one another. No other ultrastructural Kupffer cell abnormal280

THE LIVER IN MUCOPOLYSACCHARIDOSIS

(Resnick et al)

FIQURE 7. In this biopsy specimen from a patient with MPS VI hepatocellular GAG inclusions contain abundant Iipofuscln particles (straight arrow). Several lipid droplets also are present (curved arrow). (UPN 217; magnlfication x8.050.)

ities were noted. Indeed, unlike the hepatocytes, the Kupffer cells of all 27 cases showed rare or no lipid droplets. Table 3 summarizes the number and size of Kupffer cell GAG inclusions for the MPS diseases. There was no demonstrable correlation of Kupffer cell GAG inclusion number or size with specific MPS. Despite the relative paucity of hepatocellular GAG inclusions in the two MPS VI cases, the number of Kupffer cell GAG inclusions was comparable to that in the other MPS diseases. There was no correlation between patient age at time of biopsy and the Kupffer cell inclusion number or size. In the comparison of Kupffer cell with hepatocellular inclusion sizes, despite the overlapping size ranges (Tables 2 and 3), the hepatocellular inclusions generally were larger than those of Kupffer cells. Comparison of LM and EM findings revealed one discrepant case (with MPS I [UPN 13101) showing negligible colloidal-iron staining and significant GAGS ultrastructurally. Although few in number, from one to six (median, two) Ito cells (lipocytes) for each of 19 biopsy specimens (13 MPS I cases, one MPS II case, and five MPS III cases) were identified for study. In 18 of these 19 cases the Ito cells contained membrane-bound inclusions consisting of flocculent-reticular material similar in appearance to those appearing in Kupffer cells (Fig 10). Several such inclusions contained electrondense bodies, but lipofuscin particles and myelin figures were not present. The Ito cells were not remarkable for any other ultrastructural abnormalities.

Table 4 summarizes the Ito cell GAG inclusion number and size for the MPS diseases. The inclusions were more numerous in cells inflicted with MPS I than in those with MPS III; there was no correlation between inclusion size and specific disease. Furthermore, the inclusion number and size did not correspond with patient age at biopsy. There was no definite difference between the number of inclusions in the Ito and Kupffer cells. The GAG inclusions were similar in size to those of the Kupffer cells, and thus were considerably smaller than their hepatocellular counterparts. Since Ito cells could not be identified by LM with any confidence, concordance between LM and EM with regard to this cell type could not be determined. The other cellular constituents of the liver were not specifically sought. However, two cases (UPNs 977 and 1310) showed GAG inclusions in sinusoidal endothelial cells (Fig 11). The sinusoids contained many membrane-bound GAG-containing structures closely TABLE 2. Summary of Number and Size of Hepatocellular Glycosaminoglycan Inclusions by Electron Microscopy Disease MPS I MPS II MPS III MPS M

281

Median No. of Inclusions (Per ~7,450 Field) (Range) 37 43 25 0.8

(1573) (43) (E-57) (0.5-1.0)

Inclusion Diameters (cm) 8 7 10 10

x x x x

10-5-6 10-5-4 1o-5-9 1o-5-4

X x x x

1O-4 1o-4 10-4 1o-4

HUMAN

Volume 25, No. 3 (March 1994)

PATHOLOGY

FIQURE 8. The hepatocytes of both MPS VI cases contain GAG inclusions that are relatively small and scarce (arrow). (UPN 1130; original magnification ~4,608.)

resembling the hepatocellular inclusions. None of the EMS contained bile ductal epithelium. DISCUSSION In 1916 Hunter reported the first case of a peculiar syndrome of chondrodystrophic skeletal than es, heQ “Garpatosplenomegaly, and mental retardation.3’,3 goylism,” a term coined by Ellis et al in 1936,39 had been thought to result from tissue accumulation of lipid, leading to the alternative name, “lipochondrodystrophy.““j However, by 1959 the detection of increased concentrations of two GAGS (dermatan sulfate and heparan sulfate) in the urine and multiple organs justified the term “mucopolysaccharidosis.“45~343’ In 1972 Matalon 6o T l

l

10

0123456789

PatientAge

11

12

13

(Years)

FIOUREO. Scatter chart indicating the age and number of hepatocellular GAG inclusions for each MPS case (circles, MPS I: square, MPS II; diamonds, MPS Ill; triangles, MPS VI). The “bestfit” line for the 19 MPS I cases has a negative slope, showing no correlation between patlent age and inclusion number.

and Dorfman identified a-L-iduronidase as the enzyme deficient in Hurler syndrome.38 Although the clinical heterogeneity of the MPS diseases had been recognized much earlier, McKusick’s classification system was established only after the demonstration of their specific enzymatic defects.” This collection of liver biopsy specimens from 27 patients with MPS represents the largest reported series to date. Examination of the liver has allowed a qualitative and semiquantitative assessment of GAG accumulation in cells of epithelial, monocyte-macrophage, and mesenchymal derivation. The detection of excessive GAGS in both hepatocytes and Kupffer cells of all 27 patients with MPS confirms the findings ofprevious case reports and small Such GAG accumulanon was series. 5.7-8.10-12,1416,2~22.2 demonstrated by both LM and EM in all 26 biopsy specimens (for which LM slides were available), except those obtained from three patients, two showing significant hepatocellular storage material by EM only (UPNs 1040 [MPS I] and 217 [MPS VI]) and one with only ultrastructural evidence of accumulated GAGS in Kupffer cells. The lack of colloidal iron staining of hepatocytes in one MPS VI case (UPN 217) reflects the paucity of GAG inclusions observed by EM. Therefore, in reference to hepatocytes and Kupffer cells, ultrastructural examination seems to be a more sensitive means of identifying storage material than LM with the colloidaliron stain. Of six biopsy specimens without hepatocellular distension seen by hematoxylin-eosin stain, five were positive for GAGS by colloidal-iron and EM (one MPS VI patient [UPN 2171 showed few inclusions by EM and a negative colloidal-iron stain). Therefore, colloidal iron is more sensitive than the hematoxylineosin stain in detecting increased hepatocellular GAGS. The one case (UPN 1040) of hepatocellular distension seen by the hematoxylineosin stain with negative colloidaliron and positive EM remains unexplained. Since Ito 282

THE LIVER IN MUCOPOLYSACCHARIDOSIS

TABLE 4. Summary of Number and Size of Ito Cell Glycosaminoglycan Inclusions by Electron Microscopy

TABLE 3. Summary of Number and Size of Kupffer Cell Glycosaminoglycan Inclusions by Electron Microscopy Disease MPS I MPS II MPS III MPS VI

Median No. of Inclusions (Per Cell) (Range) 12 3 9 6

(622) (3) (2-14) (0.512)

Inclusion Diameters (cm)

Disease

4 4 4 8

MPS I MPS II MPS III MPS VI

x x x x

10-5-3 x 1o-5-O.9 10-5-2 x lo-L2 x

(Resnick et al)

1o-4 x 10-4 1om4 1o-4

cells could not be distinguished from Kupffer cells by LM and are fewer in number, evaluation of the Ito cells was limited to those identified by EM. Ito cells were found to contain lysosomal GAG inclusions in 18 of the 19 MPS patients for whom at least one Ito cell was identified. The accumulation of GAGS in Ito cells from patients with MPS VI has been reported in previous studies.*s4’ Although 15 of 26 biopsy specimens examined by LM showed cytoplasmic clearing of the interlobular bile ductal epithelium, none showed reactivity of these cells for colloidal iron. The following explanations may be proposed for the ductal cell cytoplasmic clearing without colloidal iron reactivity: (1) tissue processing has resulted in the washing out of accumulated GAGS, (2) the colloidal iron stain is not sufficiently sensitive to detect the amount of accumulation necessary for cytoplasmic clearing, and (3) the cytoplasmic clearing may be a paraphenomenon unrelated to the intracellular accumulation of GAGS. Since septal bile ductal cells normally contain mucin, their colloidal iron reactivity does not necessarily reflect the pathologic accumulation of GAGS. The ultrastructural features of bile ductal epithelium in MPS have yet to be characterized.

Median No. of Inclusions (Per Cell) (Range) 20 9 4 35

(10-33) (9) (O-12) (1456)

Inclusion Diameters (cm) 4 3 4 4

x x x x

10-5-10 10-5-s 1o-5-9 10-5-S

x 10-s x 10-S x 1o-5 x 10-S

By EM the GAG inclusions were distinctive and readily distinguished from those of most other lysosoma1 storage diseases.41 Nevertheless, one morphologic type of storage material in Gaucher’s disease closely resembles GAGS ultrastructurally. We attribute the presence of extracellular sinusoidal GAGS to processing artifact. All the GAG inclusion variants contained flocculent material; however, many hepatocellular inclusions also contained prominent electron-dense bodies, lipofuscin, and myelin figures. The appearance of these inclusions corresponds well to those previously shown in the literature.s~11~‘2~‘“‘6.21.25 The dense bodies appear similar to lipofuscin in most cases; however, the possibility of glycogen as the major constituent cannot be excluded. The identification of hepatocellular inclusions undergoing apparent fusion suggests that the larger lysosomal inclusions are created by coalescence of smaller ones. In accordance with Callahan and Lorincz’s EM evaluation of liver biopsy specimens from eight MPS I cases, the hepatocellular inclusions tended to be larger than those of Kupffer or Ito cells.” Most of the Kupffer and Ito cell inclusions contained the typical flocculent material, often accompanied by electrondense bodies but only infrequently showing the other

FIGURE 10. This space of Disse contains on Ito cell (orrow) harboring lipid droplets and abundant GAG inclusions. The adiocent sinusoid contains an inclusion-bearing Kupffer cell. (UPN 1104; magnification x9,780.)

283

HUMAN

PATHOLOGY

Volume 25. No. 3 (March 1994)

FIGURE 11. This sinusoid is lined by an endothelial cell (arrow) containing many GAG inclusions, several of which bear “dense bodies.” Numerous GAG inclusions float freely in the sinusoid, most likely resulting as an artifact of tissue preparation, (UPN 977; original magnification x7.640.)

aforementioned superimposed changes often seen within hepatocellular inclusions. Although examination of hepatocytes, Kupffer cells, or Ito cells is reliable for demonstrating GAG inclusions, thus distinguishing MPS cases from those of normal controls and most other lysosomal storage diseases, few ultrastructural features were restricted to a single MPS disease. However, the presence of relatively sparse small hepatocellular GAG inclusions containing abundant lipofuscin may be characteristic of MPS VI. Other investigators previously noted that the hepatocellular inclusions are less numerous and of smaller size in MPS VI than in MPS I. 15*43Since one MPS VI patient (UPN 217) underwent liver biopsy 1 month after BMT, the presence of sparse small lipofuscin-laden GAG inclusions might, in this case, have resulted from the metabolic changes associated with BMT. Also of note was the greater number of Ito cell inclusions in MPS I than in MPS III; however, this observation was based on only five MPS III cases and the study of fewer than five Ito cells per biopsy specimen. The validity of these observations will be tested as tissues from more cases of MPS III and VI are examined. The amount of storage material varied considerably among the biopsy specimens for each MPS disease. However, we were unable to confirm an earlier report of a correlation between duration of disease and amount of storage materia1.44’45 Such an assessment is limited by the heterogeneity of disease expression among patients, the lack of serial liver biopsy specimens from patients who have not undergone BMT, the small age range of our patients, and our lack of morphometric data. The variation in amount of ac-

cumulated GAGS may have resulted from allelism or differences in mutant gene expression. Although Orizaga et al reported a decreased number of hepatocellular mitochondria and increased mitochondrial electron density, we have noted no such changes.23 Since 19 of the 27 biopsy specimens were obtained from patients with Hurler syndrome, our observations hold more validity for this MPS disease than the others. Although colloidal-iron stain proved to be a more sensitive indicator of hepatocellular and Kupffer cell GAG accumulation than did hematoxylin-eosin, LM revealed pathologic changes by both histochemical techniques. By hematoxylin-eosin staining hepatocellular distension and cytoplasmic rarefaction, with obliteration of the sinusoids, were evident in 20 of the 26 biopsy specimens processed for LM. Several other case reports have made similar observations.6110*16Nevertheless, since glycogen storage disease types I and III show a similar “mosaic” pattern, LM by hematoxylineosin staining certainly is not specific for MPS.41 Furthermore, variation in fixation can alter cytoplasmic appearance, rendering the hematoxylin-eosin stain less informative. Although previous reports have described distinct hepatocellular and Kupffer cell vacuolation, such changes were inapparent among our cases. 4,611.14,1617,19,46.47 With the colloidal-iron technique the hepatocytes showed fine to coarse particulate cytoplasmic staining. The biopsy specimens generally revealed staining of most hepatocytes; however, the intensity of individual hepatocyte staining was much more variable from case to case. In contrast, the Kupffer cells, while diverse in the proportion of positive-staining cells from one biopsy specimen to another, consistently exhibited strong homogeneous individual cell reactivity. Perhaps such uniformly strong staining of the Kupffer cells resulted from the greater density of inclusions per unit area of cytoplasm in the Kupffer cells than the hepatocytes. However, as illustrated by three of our cases, LM with colloidal-iron is less sensitive than EM. Furthermore, the various patterns of colloidal-iron staining observed showed no correlation with specific MPS diseases or with duration of disease. The lack of any appreciable alcian blue staining among liver biopsy specimens from six patients indicates its insensitivity in detecting GAGS. Colloidal-iron is the stain of choice in showing the accumulation of GAGS in the MPS diseases. Few reports have described the use of colloidal-iron stain in the demonstration of GAGs.‘*‘~,~~ The ultrastructural demonstration of increased lipid vacuoles in hepatocytes of 19 patients with MPS I, MPS III, and MPS VI is of interest. In both MPS VI cases lipid inclusions were just as numerous as the GAG inclusions. The liver biopsy specimen from one MPS VI patient (UPN 217) contained 20 times as many lipid vacuoles as GAG inclusions; LM revealed frank steatosis. Perhaps the metabolic defect and accumulation of GAGS have inflicted hepatocellular injury, manifested nonspecifically as fatty change. Since only two biopsy specimens showed steatosis by LM, perhaps the ultrastructural evidence of lipid accumulation represents an early mild metabolic derangement. Interestingly, cases

284

THE LIVER IN MUCOPOLYSACCHARIDOSIS

of variable histochemical positivity of hepatocellular vacuoles for lipid were reported in the early 1950s before GAGS were recognized as the major accumulating material in these diseases5sg By EM Callahan and Lorinez reported hepatocellular lipidcontaining vacuoles accompanying the typical GAG lysosomal inclusions in liver biopsy specimens from eight MPS I patients.” One liver biopsy specimen in our study, which was obtained from a nearly 13year-old female MPS VI patient (UPN 217), showed a focal centrilobular scar. Otherwise, we found no evidence of hepatic damage. Parfrey and Hutchins reported the autopsy findings of six patients with MPS; all six livers showed extensive portal-portal bridging fibrosis.r4 These patients were 4, 16, 17, 19, 20, and 32 years of age, respectively, at the time of death-l4 Four other case reports of hepatic fibrosis in MPS appear in the older literature, ranging in severity from focal fibrosis to early cirrhosis, in patients aged 6, 6, 10, and 19 years.1g,465r Since our patients were considerably younger than those reported in the literature (median age, 1.9 years), it is possible that the development of hepatic fibrosis is correlated with duration of disease. Light microscopy with colloidal-iron stain and EM, when both applied, are sensitive in the diagnosis of MPS. Although EM is more sensitive than LM in assessing hepatocytes for GAG accumulation, the relative ease of examining a large number of Kupffer cells by LM suggests that liver biopsy specimens should be examined by both methods. Ito cells are fewer in number than Kupffer cells and thus less amenable to routine ultrastructural study. Although Haust and colleagues have advocated using alcohol fixation (or freezing) to augment metachromatic staining of GAGS with toluidine blue, colloidal-iron staining of tissue prepared by standard formalin fixation is effective and convenient.7.8 With the availability of urine GAG excretion and leukocyte enzyme assays, a cogent argument can be made against the routine use of liver biopsy for the diagnosis of MPS. Nevertheless, as illustrated by the one MPS IIIB patient in our series (UPN 659), biopsy may allow the detection of incidental cases, enabling clinicians to recognize and manage MPS early. More importantly, however, liver biopsy may be helpful in the assessment of metabolic correction after BMT, enzyme replacement, and, in the near future, gene therapy. Indeed, in our previous study we demonstrated the resolution of tissue GAG accumulation with serial liver biopsy specimens obtained from patients with MPS who had undergone BMT.’ More extensive reports of hepatic morphology in MPS, especially MPS II, MPS III, and MPS VI, are needed to define the full spectrum of changes and those features by which these syndromes might be distinguished. Efforts should be directed to confirm and explain the apparently unique hepatic morphology of MPS VI. REFERENCES 1. Neufeld EF, Muenzer J: The mucopolysaccharidoses, in &river CR, Beaudet AL, Sly WAS, et al (eds): The Metabolic Basis of

(Resnick et al)

Inherited Disease (ed 6). New York, NY, McGraw-Hill Information Services, 1989, pp 15651587 2. Krivit W, Pierpont ME, Ayaz K, et al: Bone-marrow transplantation in the Maroteaux-Lamy syndrome (mucopolysaccharidosis type VI): Biochemical and clinical status 24 months after transplantation. N Engl J Med 311:16061611,1984 3. Resnick JM, Krivit W, &over DC, et al: Pathology of the liver in mucopolysaccharidosis: Light and electron microscopic assessment before and after bone marrow transplantation. Bone Marrow Transplant 10273280, 1992 4. Bishton RL, Norman RM, Tingey A: The pathology and chemistty of a case of gargoylism. J Clin Path01 9305314, 1956 5. Brante G: Gargoylism-A mucopolysaccharidosis. Stand J Clin Lab Invest 4:43-46, 1952 6. Strauss L: The pathology of gargoylism: Report of a case and review of the literature. Am J Path01 24:855-885, 1948 7. Haust MD, Landing BH: Histochemical studies in Hurler’s dis ease: A new method for localization of acid mucopolysaccharide, and an analysis of lead acetate “fixation.” J Histochem Cytochem 9:7986, 1961 8. Haust MD, Gordon BA: Ultrastructural and biochemical aspects of the Sanfilippo syndrome-Type III genetic mucopolysaccharidosis. Connect Tissue Res 15:57-64, 1986 9. Dawson IMP: The histology and histochemistry of gargoylism. J Path01 Bacterial 67:587-604, 1954 10. Henderson JL, Macgregor AR, Thannhauser SJ, et al: The pathology and biochemistry of gargoylism. Arch Dis Child 27:230-253, 1952 11. Callahan WP, Lorincz AE: Hepatic ultrastructure in the Hurler syndrome. Am J PathoI48:277-298, 1966 12. Loeb H, Jonniaux G, Resibois A, et al: Biochemical and ultrastructural studies in Hurler’s syndrome. J Pediatr 73:860-874, 1968 13. Keller CH, BrinerJ, SchneiderJ, et al: Mukopolysaccharidose type VI-A (Morbos Maroteaux-Lamy): Korrelation der klinischen und pathologisch-anatomischen Befunde bei einem 27 jahrigen patienten. Helv Paediatr Acta 42:317-333, 1987 14. Parfrey NA, Hutchins GM: Hepatic fibrosis in the mucopolysaccharidoses. Am J Med 81:825829,1986 15. Van Hoof F: Mucopolysaccharidoses and mucolipidoses. J Clin Path01 27:6493, 1974 (suppl8) 16. Oda H, Sasaki Y, Nakatani Y, et al: Hunter’s syndrome: An ultrastructural study of an autopsy case. Acta Path01Jpn 38: 11751190,1988 17. Kressler RJ, Aegerter EE: Hurler’s syndrome (gargoylism). J Pediatr 12:579-591,1938 18. Ashby WR, Stewart RM, Watkin JH: Chondro-osteo-dystrophy of the Hurler type (gargoylism)-A pathological study. Brain 60:149179,1937 19. Kny W: Zur kenntnis der dysostosis multiplex Typ PfaundlerHurler. Z Kinderchir 63:366377, 194243 20. McKusick VA, Neufeld EF: The mucopolysaccharide storage diseases, in Stanbury JB, Wyngaarden JB, Fredrickson DS, et al (eds): The Metabolic Basis of Inherited Disease (ed 5). New York, NY, McGraw-Hill, 1983, pp 751-777 21. Van Hoof F, Hers HG: L’ultrastructure des cellules hepatiques dans la maladie de Hurler (gargoylisme). C R Acad Sci Paris 259:1281-1283, 1964 22. Callahan WI’, Hackett RL, Lorincz AE: New observations by light microscopy on liver histology in the Hurler’s syndrome. Arch Path01 83:507-512, 1967 23. Orizaga M, Haust MD, BryansAM, et al: Electron microscopic study of liver in Hurler’s syndrome. Fed Proc 22:484, 1963 (abstr) 24. Jelke H: Gargoylism. Report of a case. Ann Paediatr 177:355 373,195l 25. Murata F, Wohltman H, Spicer SS, et al: Fine structural and ultrastructural studies on the lymphocytes in three types of genetic mucopolysaccharidoses. Virchows Arch B Cell Pathol 25:61-73, 1977 26. Spicer SS, Garvin AJ, Wohltmann HJ, et al: The ultrastructure of the skin in patients with mucopolysaccharidoses. Lab Invest 31:488 502.1974 27. Spicer SS, Garvin AJ, Simson JAV, et al: Cytochemistry of the skin of patients with mucopolysaccharidoses. Histochem J 10:137-150, 1978 28. Whitley CB, Gorlin RJ, Krivit W: A nonpathologic allele (Iv”) for low cr-L-iduronidase enzyme activity visa-+ prenatal diagnosis of Hurler syndrome. Am J Med Genet 28:233243,1987

285

HUMAN

PATHOLOGY

Volume 25, No. 3 (March 1994)

29. DiFerrante N, Neri G, Neri ME, et al: Measurement of urinary glycosaminoglycans with quaternary ammonium salts: Extension of the method. Connect Tissue Res 1:93-101, 1972 30. Luna LG: Methods for carbohydrates and mucoproteins, in Manual of Histologic Staining Methods of the AFIP (ed 3). Washington, DC, McGraw-Hill, 1968, pp 163168 31. Hunter C: A rare disease in two brothers. Proc R Sot Med 10:104116,191617 32. Hurler G: Uber einen Typ multipler Abartungen, vorwiegend am Skelettsystem. Z Kinderchir 24:220-234, 1919-20 33. Ellis RWB, Sheldon W, Capon NB: Gargoylism (chondro-osteodystophy, cornea1 opacities, hepatosplenomegaly, and mental deficiency). Q J Med 29:119-l 39, 1936 34. Brown DH: Tissue storage of mucopolysaccharidosis in Hiirler-Pfaundler’s disease. Proc Nad Acad Sci USA 43:783790, 1957 35. Dorfman A, Lorincz AE: Occurrence of urinary acid mucopolysaccharides in the Hurler syndrome. Proc Nat1 Acad Sci USA 43:443-446,1957 36. Meyer K, Grumhach MM, Linker A, et al: Excretion of sulfated mucopolysaccharides in gargoylism (Hurler’s syndrome). Proc Sot Exp Biol Med 97:275-279, 1958 37. Meyer K, Hoffman P, Linker A, et al: Sulfated mucopolysaccharides of urine and organs in gargoylism (Hurler’s syndrome). II. Proc Sot Exp Biol Med 102:587-590,1959 38. Matalon R, Dorfman A: Hurler’s syndrome, an a-L-iduronidase deficiency. Biochem Biophys Res Commun 47:95%964,1972 39. McKusick VA: Genetic nosology: Three approaches. Am J Hum Genet 30: 105-l 11, 1978 40. McGovern MM, Ludman MD, Short MP, et al: Bone marrow transplantation in Maroteaux-Lamy syndrome (MPS type 6) : Status 40 months after BMT, in Krivit W, Paul NW (eds): Bone Marrow Trans-

plantation for Treatment of Lysosomal Storage Diseases. Birth Defects: Original Article Series XXII. New York, NY, Liss, 1986, pp 41-53 41. Snover DC: Metabolic diseases, in Snover DC (ed): Biopsy Diagnosis of Liver Disease. Baltimore, MD, Williams & Wilkins, 1992, pp 52-87 42. Phillips JM, Poucell S, Patterson J, et al: Metabolic liver disease, in The Liver: An Atlas and Text of Ultrastructural Pathology. New York, NY, Raven, 1987, pp 248,250,304305,314 43. Tondeur M, Neufeld EF: The mucopolysaccharidoses: Biochemistry and ultrastructure, in Good RA, Day SB, Yunis JJ (eds): Molecular Pathology. Springfield, MA, Charles C. Thomas, 1975, pp 600-62 1 44. Haust MD, Gordon BA, Hong R, et al: Clinicopathological conference: An adolescent girl with severe mental impairment and mucopolysacchariduria. Am J Med Genet 22:1-27, 1985 45. Haust MD, Gordon BA, Bryans AM, et al: Heparitin sulfate mucopolysaccharidosis (Sanfilippo disease): A case study with ultrastructural, biochemical, and radiological findings. Pediatr Res 5:137150,197l 46. De Lange C: Over dysostosis multiplex typhus Hurler of Gargoylismus. Psychiatr Neurol Bladen 46:2-18, 1942 47. Zeeman WPC: Gargoylismus. Acta Ophthalmol (Copenh) 20:4@47, 1942 48. Reilly WA: An atypical familial endocdnopathy in males with a syndrome of other defects. Endocrinology 19:639-648, 1935 49. De Lange C, Gerlings PG. de Kleyn A, et al: Some remarks on gargoylism. Acta Paediatr 31:398416, 194344 50. Lindsay S, Reilly WA, Gotham TJ, et al: Gargoylism. II. Study of pathologic lesions and clinical review of twelve cases. Am J Dis Child 76:239-306, 1948 51. Schwarz H, Gagne R: A case of gargoylism. Can Med Assoc J 66:375-377,1952

286