JOURNAL OF BONE AND MtNERAL RESEARCH Volume 6, Number 3, 1991 Mary Ann tiebert, tnc., Publishers
Iron as a Possible Aggravating Factor for Osteopathy in Itai-itai Disease, a Disease Associated with Chronic Cadmium Intoxication MAKOTO NODA, MASANORI YASUDA, and MASANOBU KITAGAWA
ABSTRACT
Ifai-itai disease is thought to be the result of chronic cadmium (Cd) intoxication. We examined 23 autopsy cases of itai-itai disease and 18 cases of sudden death as controls. Urine and blood samples from 10 patients were collected before they died and revealed the presence of severe anemia and renal tubular injuries. Undecalcified sections of iliac bone were stained with Aluminon reagent, an ammonium salt of aurintricarboxylic acid, and Prussian blue reagent in all cases of itai-itai disease. These two reagents reacted at the same mineralization fronts. X-ray microanalysis revealed the presence of iron at mineralization fronts in itai-itai disease. Five patients showed evidence of hemosiderosis in the liver, spleen, and pancreas, probably as a result of post transfusion iron overload. Renal calculi and calcified aortic walls were also stained with Prussian blue reagent in several patients. Neither ferritin nor transferrin were visualized at mineralization fronts in itai-itai disease by immunohistochemical staining. These results suggest that iron is bound to calcium or to calcium phosphate by a physicochemical reaction. A marked osteomalacia was observed in 10 cases of itaiitai disease by histomorphometry. Regression analyses of data from cases of itai-itai disease suggested that an Aluminon-positive metal inhibited mineralization and that renal tubules were injured. Since bone Cd levels were increased in itai-itai disease, it is likely that renal tubules were injured by exposure to Cd. Therefore, stainable bone iron is another possible aggravating factor for osteopathy in itai-itai disease, and a synergistic effect between iron and Cd on mineralization is proposed.
INTRODUCTION sometimes have serious deleterious effects on bone systems.") h i - i f u i ("ouch-ouch'' in English) disease (IID) is thought to be the result of chronic cadmium (Cd) intoxication, and patients suffer from severe osteomalacia or osteoporosis, or both, with high frequency.") It is well known that idiopathic hemochromatosis and 6-thalassemia are associated with osteopathy, and posttransfusion iron overload has been suggested as a cause of Aluminum (Al) was found in patients with dialysis-associated osteomalacia as a toxic agent for osteopathy. ( 5 ) Recently, Pierided6)suggested that iron can also be found in some patients with dialysis-associated osteomalacia. In a previous report we described how both osteoblast
T
RACE METALS
function and mineralization were impaired in patients with IID, as indicated by quantitative bone histology, and suggested that Cd may be one of the causative It is not generally recognized that undecalcified bone sections from patients with IID react with Aluminon reagent, an ammonium salt of aurintricarboxylic acid, at mineralization Aluminon reagent reacts primarily with tissue Al,'8.9)but this reagent also cross-reacts with tissue iron and other metals.(8~1a) In this study we showed the presence of stainable bone iron in 23 cases of IID, and we suggested that iron is a possible aggravating factor for osteopathy in IID.
MATERIALS AND METHODS
Selection of patients A total of 23 autopsy cases of IID, patients aged 62-84
Department of Pathology, Toyama Medical and Pharmaceutical University, Faculty of Medicine, Toyama, Japan.
245
NODA ET AL.
246 years (mean + SD, 76.6 f 5 . 9 , were examined: 1 man was aged 80 years, and 22 women were aged 62-84 years (76.4 f 5.5 years). The clinical diagnosis is described in our previous The causes of death were bronchopneumonia (7 cases), hypovolemic shock due to gastroduodenal ulcer (5 cases), lung cancer (1 case), pulmonary abscess (1 case), uremia (1 case), sepsis (1 case), urinary tract infection (3 cases), colitis (1 case), and asphyxia (3 cases). For the controls, 18 autopsy cases of women aged 60-91 years (75.7 f 8.3 years) were selected. Each had died suddenly, and causes of death included rupture of cerebral aneurysm (3 cases), burn (1 case), accident (4 cases), rupture of dissecting aneurysm ( 1 case), thrombosis of superior mesenteric artery ( 2 cases), and asphyxia (7 cases). None had a history of prolonged bed rest, immobilization, or any organic, endocrine, or metabolic disease. No glucocorticoids, antacids, or diuretics had been prescribed except at the terminal stage. Blood and urine samples were collected, before they died, from 10 of the 23 cases of IID. These 10 were all women aged 62-84 years (75.1 f 6.4). Serum concentrations of calcium, inorganic phosphorus, alkaline phosphatase, and creatinine and urinary concentrations of inorganic phosphorus and creatinine were measured by standard automated methods. Urinary fi,-microglobulin was measured by a single radial immunodiffusion method.'") Urinary lysozyme was measured by the lysoplate method."z' Urinary and bone levels of Cd were determined by flameless atomic absorption spectrophotometry after extraction with ammonium pyrolidine dithiocarbamate (APDC) and methyl isobutyl ketone (MIBK).'") Control values of urinary and serum chemistries were taken from a report by S h i g e m a t s ~ , (who ~ ~ ) summarized extensive epidemiologic and clinical investigations of inhabitants of Cd-polluted and unpolluted areas. Bone levels of Cd for controls were also taken from our previous report. ( ' ) All necropsy materials were taken from a standard site, approximately 2 cm posterior and inferior to the left anterior-superior iliac spine.
Histochemistry The autopsy specimens of iliac bone were fixed in 70% ethanol, embedded in methyl methacrylate, and sectioned in 5 pm thick slices with a Jung K microtome. The autopsy specimens were also fixed in 15% buffered formalin and postfixed in a 0.5% solution of cyanuric chloride that contained 1% N-methylmorpholin for 1 day. They were then decalcified with 10% ethylenediaminetetraacetic acid in distilled water, embedded in paraffin, and sectioned into slices of less than 3 pm thickness [Yoshiki's method(14)]. Undecalcified and decalcified sections of bone were stained with Prussian blue for iron(15)and with Aluminon reagent(8-Lo)for Al and/or iron. Undecalcified sections of bone were double-stained, first with Aluminon reagent and then with Prussian blue reagent, to examine the localization of the metal reaction. Undecalcified sections of bone were stained with Villanueva's solution.(I6)Decalcified sec-
tions of bone were stained with hematoxylin and eosin according to the Yoshiki method.(*4) The kidney, liver, spleen, pancreas, and calcified aortic walls of the autopsy materials were fixed in 15% buffered formalin, embedded in paraffin, sectioned in 3 pm thick slices, and stained with hematoxylin and eosin and Prussian blue reagent.
X-ray microanalysis Undecalcified, methyl methacrylate-embedded sections of bone from patients and control subjects were placed on carbon plates, rinsed with distilled water, and allowed to dry before being coated with carbon. X-ray microanalysis (XMA) combined with scanning electron microscopy (SEM) was carried out for examination of trabecular bone using a Hitachi X-650electron microscope equipped with an energy-dispersive XMA system Kevex 7000 operated at 20 kV. We used point XMA for the detection of iron in the bone by counting x-ray pulses from all elements in a given spot. We also set a window from 6.26 to 6.64 keV and obtained iron La (6.40 keV) x-ray pulses for line XMA, which counts iron-specific pulses of x-rays along a scanning line.
Immunohisrochemistry Immunohistochemical staining was applied to samples from 23 cases of IID using the avidin-biotin-peroxidase complex (ABC) m e t h ~ d . " ~Decalcified ) sections of bone were deparaffinized in xylol and pretreated with 1% hydrogen peroxide (H,O,) for 30 minutes. They were incubated with normal goat serum for polyclonal antibodies for 20 minutes. Sections were reacted with polyclonal antibodies raised in rabbits against transferrin and ferritin (Dako Corporation, USA) for 30 minutes. Next they were treated with biotinylated goat serum raised against rabbit immunoglobulin and then reacted with ABC (Vector Laboratories, Burlingame, CA) for 30 minutes. All peroxidase reactions were localized by reaction with 3,3'-diaminobenzidine tetrahydrochloride (Wako, Japan) that contained 0.001 9'0 H,O,. Sections were counterstained with methyl green. The specificity of immunoreactivity was confirmed by replacing the antibody with control rabbit serum.
Histomorphometry Eroded surface, quiescent surface, osteoblasts, and osteoclasts have been defined and described elsewhere.(Is) The entire area of cancellous tissue in the bone sections was quantitatively analyzed by a semiautomatic rnethod,l7) and at least 60 optical fields were analyzed. Histomorphometric measurements are described with reference to the nomenclature of Parfitt and colleagues,''8) as follows: ( I ) bone volume, the fraction of tissue volume occupied by trabecular bone (BV/TV); ( 2 ) osteoid volume, the fraction of tissue volume occupied by osteoid (OVITV); (3) mineralized volume, the fraction of tissue volume occupied by mineralized bone (Md.V/TV); (4) wall thickness, the mean
IRON IN ITAI-ITAI DISEASE distance between cement lines and the trabecular surfaces of completed structural units (W.Th); (5) osteoid thickness, the mean distance between mineralization fronts and osteoid surfaces (0.Th); (6) osteoblast surface, the fraction of osteoid surface covered by osteoblasts (Ob.S/OS); (7)osteoid surface, the fraction of trabecular surface covered by osteoid (OS/BS); (8)osteoclast surface, the fraction of eroded surface covered by osteoclasts (Oc.S/ES); (9)eroded surface, the fraction of trabecular surface occupied by eroded surface (ES/BS); (10) Aluminon-stainable bone, the fraction of trabecular surface covered with Aluminon (Aluminon/BS); and (1 1) Aluminon-stainable bone (osteoid referent), the fraction of osteoid surface stained with Aluminon (Aluminon/OS). The statistical significance of differences was determined by Student's t-test for the two group means of IID and control subjects. A covariance analysis with regression was applied to laboratory and histomorphometric data to determine correlation coefficients.
247 centrations of 1,25-dihydroxycholecalciferol [ 1,25(OH),D,] in this study. These results showed that cases of IID were associated with the presence of severe anemia and damage to renal tubules.
Histochemistry
Figures 1 and 2 show the results of staining with Aluminon and Prussian bue. Both reagents reacted at mineralization fronts in undecalcified sections of bone from patients with IID. These two reagents occasionally reacted on osteoid-free surfaces of trabecular bone. Since Aluminon reagent reacts with tissue Al as a reddish line and Prussian blue reagent reacts with tissue iron as a bluish line, it is possible to discriminate Al from iron if appropriate double-staining is performed on the same section. A clear, reddish line was observed in a case of IID at a mineralization front stained with Aluminon reagent (Fig. 3A), and a bluish line overlapped at the identical front after staining the same section with Prussian blue reagent (Fig. 3B). Similar results were obtained by double-staining with Aluminon and Prussian blue in undecalcified sections of bone from RESULTS 23 patients with IID. These results suggest the possibility Urinary and serum chemistry that Aluminon reagent reacts with tissue iron at mineralThe results of the analysis of urinary and serum chemis- ization fronts of undecalcified sections of bone in IID. Detries are listed in Table 1. We did not apply statistical treat- calcified sections of bone from the patients with 11D did ment to the analysis of the biochemical data from the pa- not react with Aluminon or with Prussian blue reagents. tients and control subjects. However, there were marked Decalcified and undecalcified sections of bone from the decreases in hemoglobin, hematocrit, percentage tubular control subjects did not react with either of these two rereabsorption of phosphate, and increases in urinary P2-mi- agents. Excess tissue hemosiderin was observed in the liver, croglobulin, urinary lysozyme, serum alkaline phosphatase, and serum creatinine. We observed the increase in spleen, and pancreas from five patients with IID after bone Cd content in IID, but we did not measure the net Prussian blue staining. No signs of idiopathic hemochrocontent of Al in this study. We did not measure the con- matosis were confirmed clinically or histologically. We
TABLE 1. URINARY A N D SERUM CHEMISTRY IN 10 PATIENTS WITH
Variable Hemoglobin, g/dl Hematocrit, 070 Urinary &-microglobulin, mg/dl Urinary lysozyme, mg/dl Urinary creatinine, mg/dl Urinary inorganic phosphorus, g/dl Urinary cadmium content, pg/liter Tubular reabsorption of phosphate, 070 Serum alkaline phosphatase, BLU/liter Serum calcium, mg/dl Serum inorganic phosphorus, rng/dl Serum creatinine, mg/dl Bone cadmium content, p g / g wet weight
Controlb
13.3 f 38.9 * 0.18 f 0.71 * 36.4 f 19.1 f 5.47 f 84.1 + 2.76 f 9.76 f 3.64 f 0.93 f 0.50
f
1.5 3.9 0.45 1.92 25.7 16.4 3.84 5.2 0.78 0.36 0.47 0.28 0.50
ITAI-ITAI
DISEASEa
Itai-itai disease (n
= 10)
6.5 f 0.5 20.3 f 2.7 3.63 + 1.23 6.29 + 3.67 31.2 f 7.0 16.2 f 5.0 4.98 + 2.03 27.2 f 8.4 5.06 f 2.09 8.37 f 0.96 3.74 + 1.16 4.69 f 1.55 2.24 f 0.87
aMean f SD. hControl values for the parameters of urinary and serum chemistry are taken from Shigemat~u,"~' and bone cadmium contents are taken from our previous report"' with the help of Dr. Kuzuhara from the National Institute of Public Health.
248
NODA ET AL.
I
FIG. 1. Photomicrograph of an undecalcified section of bone from a patient with itai-itai disease stained with Aluminon reagent. Note clear (reddish) lines located at the osteoid/bone interface. Aluminon stain, x 20 (original magnification).
2#
FIG. 2. Photomicrograph of an undecalcified section of bone from a patient with itai-itai disease stained with Prussian blue reagent. Note clear (bluish) lines located at the osteoid/bone interface as in Fig. 1. Prussian blue stain, x 20 (original magnification).
could not clarify the exact amount of iron loaded by therapeutic measure in this retrospective study, and thus we could not clarify dose-response relationships between the total amount of loaded iron and the extent of bone lesions. However, there is a possibility that some of the patients received red blood transfusions or parenteral iron or both
for severe anemia. Hemosiderin-laden erythroblasts in bone marrow always reacted with Prussian blue reagent, irrespective of whether bone sections were decalcified. Erythroblasts did not react with Aluminon reagent in decalcified or undecalcified sections of bone. Mild nephrocalcinosis was observed in several cases of
249
IRON IN ITAI-ITAI DISEASE
A
-
B g X‘
+ FIG. 3. Photomicrographs of an undecalcified section of bone from a patient with itai-itai disease stained with both Aluminon and Prussian blue reagents. (A) Aluminon staining only; and (B) Prussian blue staining after the Aluminon staining shown in A. Note the overlap between the reddish line (Aluminon stain) and the bluish line (Prussian blue stain) at mineralization fronts. Aluminon and Prussian blue stain, x 100 (original magnification).
IID, probably due to excessive dietary intake of vitamin D administration for treatment of severe osteoporosis. These renal calculi were stained with Prussian blue reagent (Fig. 4). Calcified aortic walls were also stained with Prussian blue reagent (Fig. 5 ) . These results suggest that iron is bound to calcium or to calcium phosphate at mineralization fronts, in renal calculi, in aortic walls, and presurnably in other calcified tissues of patients with 11D.
X-ray microanalysis Figure 6 shows a scanning electron micrograph of trabecular bone from a patient with IID. XMA revealed that undecalcified sections of bone contained unusual amounts of iron (Fig. 7). This result was reproduced in all cases of IID. Aluminum and Cd were not detected in any cases of IID at a significant level by this technique. Careful point XMA revealed that the counts due to iron were highest at mineralization fronts. These results were supported by the results of line XMA, which produced clear biphasic peaks (Fig. 6 ) . These peaks coincided with the distributions of
iron at mineralization fronts where both Prussian blue and Aluminon reagents reacted. Iron, Al, and Cd were not detected in control subjects at any significant level by XMA. These results suggest that undecalcified sections of bone from patients with IID contain iron that is deposited predominantly at mineralization fronts.
Immunohistochemistry Immunohistochernical staining revealed that ferritin-specific and transferrin-specific antibodies failed to react at mineralization fronts but reacted with hernatopoietic cells in bone marrow. These results suggest that trabecular bone in IID does not contain ferritin or transferrin at mineralization fronts. These results also suggest that iron is bound to calcium or calcium phosphate at mineralization fronts as “free” ions.
Histomorphometry Table 2 shows the results of histomorphometric data from 10 cases of IID and 18 control subjects. In the case of
250
NODA ET AL.
FIG. 4. Photomicrograph of a renal section from a patient with itai-itai disease stained with Prussian blue. Renal calculi are stained with Prussian blue reagent and visible as fine granular deposits. Prussian blue stain, x 50 (original magnification).
FIG. 5. Photomicrograph of a calcified aortic wall from a patient with itai-itai disease stained with Prussian blue stain. Calcified tissue are stained with Prussian blue reagent and visible as fine granular deposits. Prussian blue stain, x 50 (original magnification).
all parameters of bone structure and of the formation and resorption of bone, significant differences were found between group means for the patients and control subjects. Osteoid volume, osteoid thickness, osteoblast surface, 0steoid surface, and eroded surface were significantly greater in patients (p < 0.01). Mineralized volume and wall thick-
ness were significantly lower (p < 0.01). The significant increase in bone volume (p < 0.01) was due to an increase in osteoid volume. These results suggest that patients with IID experience a marked osteomalacia with or without reduction in bone mass. The results of regression analyses of data from 10 cases
25 1
IRON IN ITAI-ITAI DISEASE
FIG. 6. Scanning electron micrograph of trabecular bone with excess osteoid accumulation from a patient with itaiitai disease. The upper straight line is a scanning line. The lower zigzag line indicates the result of line XMA, which counts iron-specific pulses of x-rays along a scanning line. Two peaks (arrows) are seen at the mineralization fronts of the trabecular bone. Asterisk (*) indicates the spot at which point XMA was carried out. x 1OOO.
that the area of Aluminon-stained surface with osteoid referent was inversely correlated with osteoid volume ( r = -0.504) and osteoblast surface ( r = -0.490). The bone Cd content was inversely correlated with urinary inorganic phosphorus ( r = -0.624, p < 0.05). Table 4 shows a list of coefficients of correlation between histomorphometric data and laboratory data. Urinary lysozyme increases with osteoid volume ( r = 0.738, p < 0.05), and serum inorganic phosphorus also increases with both osteoid volume ( r = 0.767, p < 0.01) and osteoid thickness ( r = 0.683, p < 0.05). These results suggest that if renal tubules are injured and, thus, 1,25-(OH),D, is not produced adequately because of injury to renal tubules, reabsorption of lysozyme, excretion of inorganic phosphorus, and bone mineralization are impaired. If the percentage tubular reabsorption of phosphate decreases, as is frequently found in IID, the osteoblast surface increases ( r = -0.656, p < 0.05). Since Cd usually inhibits osteoblast f u n ~ t i o n , ~ ~this ~ ' ' 'result suggests the possibility that Cd, if present, may affect renal tubules directly. Osteoclast surface was correlated positively both with serum creatinine ( r = 0.794, p < 0.01) and with serum alkaline phosphatase ( r = 0.488), suggesting that the increase in osteoclast surface may relate to the advances in renal insufficiency. Osteoclast surface was also correlated positively with Aluminon-stained surface, as shown in Table 3 ( r = 0.651, p < 0.05). Aluminon-positive metal may affect the progress of renal insufficiency or bone resorption or both, although we did not measure concentrations of serum parathyroid hormone in this study. The present histomorphometric data indicate that iron may impair bone mineralization and osteoblast function in cases of IID with osteomalacia. Furthermore, Cd may injure renal tubules directly.
DISCUSSION
This study showed the presence of iron at mineralization fronts in undecalcified sections of bone from patients with IID. This result was clearly demonstrated by staining with Prussian blue, which reacts with tissue iron with a high deof IID are shown in Tables 3 and 4. Table 3 shows a list of gree of sensitivity and sepcificity,"') and confirmed by correlation coefficients between Aluminon- or Cd-related point and line XMA. Undecalcified sections of bone also data and histomorphometric and laboratory data. Alumi- reacted with Aluminon reagent in all cases with 1ID. Alunon-stained surface was correlated inversely with mineral- minon reagent reacts primarily with Al, and Al is thought ized volume ( r = -0.743, p < 0.05) and positively with to be a causative metal for osteomalacia associated with osteoid surface ( r = 0.740, p < 0.05), suggesting that Alu- dialysis." ' Aluminon reagent also cross-reacts with tisminon-positive metal inhibited mineralization. This sug- sue iron.'8 l o ) We observed that both Aluminon and Prusgestion is supported by the result that Aluminon-stained sian blue reagents reacted at the same mineralization fronts surface area was correlated inversely with wall thickness ( r by double-staining of undecalcified section of bone. We = -0.501) and bone volume (r = -0.495). Aluminonalso observed that bone Cd content in IID was increased in stained surface with osteoid referent defines the fraction of this study and in our previous report,'7) although Cd and osteoid surface that is covered with Aluminon, and this pa- Al were not detected by XMA at significant levels in this rameter was correlated inversely with osteoid thickness (r study. It therefore appears that bones of patients with IID = -0.682, p < 0.05). Thus if the area of Aluminon- contain iron as well as Cd, and that iron, like Al in dialysis stained surface increases, then osteoid thickness decreases osteomalacia, may play important roles in osteopathy. within an estimated osteoid area. This result suggests that Idiopathic heniochromatosis rarely complicates osteoAluminon-positive metal inhibits the formation of osteoid penia.I3) De Vernejoul et al.'" reported a hyperosteoidosis by osteoblasts. This suggestion is supported by the result and mineralization defect in patients with 0-thalassemia.
252
NODA ET AL.
FIG. 7. Spectra of point XMA of the trabecular bone seen in Fig. 6 at the spot marked *. The spectrum in A shows a wide range of energies, and the spectrum in B shows a narrow range of the energies shown in A. Note the presence of iron and the absence of Cd and Al.
TABLE 2. HISTOMORPHOMETRIC DATAFOR 10 PATIENTS WITH ITAI-ITAIDISEASE^ ~~
Variable
Control(n
Bone volume, To Osteoid volume, Vo Mineralized bone volume, 070 Wall thickness, pm Osteoid thickness, pm Osteoblast surface, 070 Osteoid surface, Vo Osteoclast surface, Vo Eroded surface, Vo Aluminon-stainable surface, Vo Aluminon-stainable surface, To (osteoid referent)
10.3 0.10 10.2 51.4 10.0 3.10 10.1 8.40 8.31
=
18)
f 4.8 f
f f f
= f
+ f
0.08 4.8 7.2 2.1 4.49 7.4 9.10 7.28
Not stained Not stained
(n
=
IO)
f 6.8b 5.92 + 5.19b 5.8 f 5.Ob 46.1 f 4.9b 41.3 + 19.8b 3.84 f 3.37b 83.4 f 9.8b 23.07 + 13.39b 9.04 f 3.93b 43.1 + 10.0 55.9 f 12.2
11.7
aMean f SD. bp < 0.01.
Recently Pieridesc6)reported the association between iron overload and osteomalacia in hemodialysis patients. Phelps et al.(”] found tissue iron but not A1 in a case of dialysis osteomalacia. These authors showed iron at mineralization fronts by Prussian blue staining and suggested that the iron was brought about by overdose of dialysate or red
blood transfusions. Severe anemia was present in our patients. Five patients also showed evidence of posttransfusion hemosiderosis in the liver, spleen, and pancreas. These patients may have received an overdose of exogenous iron supplements. Many patients suffer from severe anemia due to various
253
IRON IN ITAI-ITAI DISEASE
TABLE 3. CORRELATION COEFFICIENTS BETWEENLABORATORY AND HISTOMORPHOMETRIC DATAAND ALUMINONAND CADMIUM-RELATED DATAIN 10 CASESOF ITAI-ITAIDISEASE
Aluminon-stainable surface, Yo Aluminon-stainable surface, Yo (osteoid referent) Urinary Cd content
Bone Cd content
ar,
Wall thickness, -0.501 Bone volume, -0.495
Mineralized volume, -0.743b Osteoid surface, 0.740b Osteoclast surface, 0.651b Osteoid thickness, -0.682b Bone volume, -0.606b
Urinary creatinine, -0.705h Urinary inorganic phosphorus, -0.624b
Osteoid volume, -0.504 Osteoblast surface, -0.490 Serum calcium, 0.402 Osteoid volume, -0.533 Urinary lysozyme, -0.496 Wall thickness, -0.460 Serum inorganic phosphorus, -0.457 Osteoclast surface, -0.525
coefficient of correlation.
bp < 0.05.
TABLE4. CORRELATION COEFFICIENTS BETWEEN HISTOMORPHOMETRIC DATAAND LABORATORY DATA IN 10 CASESOF ITAI-ITAI DISEASE
Urinary &-microglobulin Urinary lysozyme
Osteoid volume, 0.738b
Urinary creatinine Urinary inorganic phosphorus To Tubular reabsorption of phosphate
Eroded surface, -0.636b Osteoblast surface, -0.656h
Serum alkaline phosphatase Serum inorganic phosphorus
Eroded surface, 0.863~ Osteoid volume, 0.767~ Osteoid thickness, 0.683b Osteoid thickness, -0.614b Osteoclast surface, 0.794" Osteoblast surface, 0.655b
Serum calcium Serum creatinine
ar,
Osteoblast surface, 0.437 Eroded surface, -0.419 Bone volume, 0.549 Osteoid thickness, 0.536 Wall thickness, 0.47 1 Osteoclast surface, 0.441 Mineralized volume, -0.575 Osteoid thickness, -0.486 Osteoclast surface, 0.488 Osteoid surface, 0.493 Osteoclast surface, 0.432 Osteoid volume, -0.556 Eroded surface, 0.489 Osteoid thickness, 0.472 Mineralized volume, -0.441
coefficient of correlation.
bp c: 0.05. ' p < 0.01.
underlying diseases and receive repeated red blood transfusions and/or parenteral iron, but the frequency of iron-induced osteopathy is very low. There were no patients in our study who had undergone hemodialysis. Some patients with IID had severe osteomalacia and osteoporosis, but they did not reveal clinical signs of anemia or past history of red blood transfusions. ' z ) Concentrations of iron were reported to be normal or even low in the liver, pancreas, and kidney of people living in Cd-polluted areas.'22)
Therefore, these observations d o not necessarily point to iron as a single etiologic agent for osteopathy in IID. Body iron is usually bound by or incorporated into various proteins, for example hemoglobin iron, storage iron, and transport iron, and it does not exist as a free ati ion."^' If iron-deficiency anemia is caused by malnutrition, multiple pregnancies, and other factors, as in IID, and if exogenous iron is not supplied, it is unlikely that body iron shifts to bone as a specific storage site. How-
NODA ET AL.
254
ever, in this study stainable bone iron was clearly observed in all cases examined. The reason for stainable bone iron in iron-deficient patients with IID is currently unknown, but it may be explained in part by the damage to tissue caused by Cd. Itai-itai disease was observed in inhabitants of Cdpolluted areas, and we showed an increase in bone Cd content in patients with IID. There is a possibility that patients in this study were exposed to Cd. The damage to renal tubules was demonstrated by the results of urinary and serum chemistries and of histomorphometry. Since serum chemistries showed advanced renal insufficiency in IID, the decrease in the percentage tubular reabsorption of phosphate alone may explain the increased phosphaturia per nephron and does not necessarily indicate the presence of injury t o renal tubules. We did not measure phosphorus excretion in patients exposed to Cd before they develop chronic renal insufficiency. However, an injury to renal tubules is a frequent finding in clinical and experimental studies of Cd intoxication.I2’ Therefore, renal tubules in IID seemed t o be injured by exposure to Cd, and tissue damage caused by Cd may relate to the onset and progress of iron accumulation in bone. Cadmium appears to compete with iron at the transfer system in the intestinal wall, and absorption of Cd increases when iron-deficiency anemia is present.(z4’Christoffersen et aI.(”) reported that Cd ions were adsorbed onto crystals of calcium hydroxyapatite and incorporated into the crystals, making them very resistant to subsequent dissolution. Some investigators have reported that iron ions can be incorporated in hydroxyapatite crystals or absorbed on the surface of teeth and bone^.^^^^"^ These reports suggest that both C d and iron ions affect the growth of crystals of calcium hydroxyapatite and contribute to the pathologic changes in the bone tissue. Both Cd and iron belong to a group of multivalent heavy metals and are bound to plasma transferrin with the same stoichiometry.‘z8.29) Meyer et al.‘Jo)reported that there is a synergistic effect between trivalent iron and citric acid on the growth of crystals of calcium phosphate, and that an iron-citrate complex may inhibit biologic calcification. Therefore, the synergistic effect not only of iron but also of Cd on the growth of crystals of calcium hydroxyapatite is more likely to impair bone mineralization than the effect of each metal alone. The results from histochemistry and immunohistochemistry suggest that iron is present as “free” ions and is bound to calcium or calcium phosphate as the result of a physicochemical reaction. Recently, Laeng et a1.(31)postulated the presence of an osteoid pool of iron in patients with primary hemochromatosis as an unusual site of iron storage. It is well known that long-term administration of Cd produces bone marrow hypoplasia and iron deficiency and thus decreases tissue utilization of Heubers et al.1331reported the effect of Cd on the release of iron from transferrin in iron-deficient rats. Therefore it is possible that body iron could shift to the bone as the result of a physicochemical reaction and that bone would then serve as an unusual site for storage of iron, when tissue utilization of iron is altered by pre-existing exposure to Cd and/ or other toxic metals. If this hypothesis is valid, long-term
therapy with iron supplements and blood transfusions for severe anemia and administration of 1 ,25-(OH),D3, which induces iron uptake,1341as a treatment for osteoporosis may modify the clinical course of IID. Alternatively, desferrioxamine may be of great value for the removal of iron and other metals in this d i s e a ~ e . ‘ ~ ~ . ~ ~ ) In conclusion, stainable bone iron was clearly demonstrated by histochemistry and XMA, and injuries to renal tubules were suggested in all cases with 11D. The stainable bone iron is a possible aggravating factor for osteopathy in IID, and a synergistic effect between iron and Cd on bone mineralization seems likely.
ACKNOWLEDGMENTS This study was performed in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Ph.D.) at Toyama Medical and Pharmaceutical University. We thank Drs. A. Miwa and M. Murai and Mr. H . Yamane for their efforts in subject sampling and their encouragement. We are grateful to Dr. Y. Kuzuhara of the National Institute of Public Health for contributing the measurements of the levels of Cd in bone. We express our thanks to Mr. T. Kumada and Mr. M. Kawahara for excellent technical assistance. We also thank Mr. T. Nagata and Mr. E. Tahara for preparation of tissues.
REFERENCES 1. Priest ND (ed.) 1985 Metals in Bone. M T P Press Limited,
Lancaster. 2. Nogawa K 1981 Itai-itai disease and follow-up studies. In: Nriagu J O (ed.) Cadmium in the Environment. John Wiley & Sons, New York, pp. 1-37. 3. Delbarre F 1960 L’Osteoporose de I’hemochromatose. Semin Hop Paris 36:3279-3284. 4. De Vernejoul MC, Girot R, Gueris J , Cancela L, Bang S, Bielakoff J , Mautalen C, Goldberg D, Miravet L 1982 Cal-
cium phosphate metabolism and bone disease in patients with hornozygous thalassemia. J Clin Endocrinol Metab 54:276281. 5 . Ellis HA, McCarthy J H , Herrington J 1979 Bone aluminum
6. 7.
8. 9.
in haernodialysed patients and in rats injected with aluminum chloride: Relationship to impaired mineralization. J Clin Pathol 32:832-839. Pierides AM 1984 Iron and aluminum osteomalacia in hemodialysis patients (letter). N Engl J Med 310:323. Noda M, Kitagawa M 1990 A quantitative study of iliac bone histopathology on 62 cases with itai-itai disease. Calcif Tissue Int 47:M-74. Hammett LP, Sottery C T 1925 A new reagent for aluminum. J Am Chem SOC47:142-143. Maloney NA, Ott SM, Alfrey AC, Miller NL, Coburn JW, Sherrard DJ 1982 Histological quantitation of aluminum in iliac bone from patients with renal failure. J Lab Clin Med 99:206-216.
10. Savory J, Wills MR 1986 Analytical methods for aluminum measurement. Kidney Int 29524-27. 11. Crowle AJ 1988 Detection and measurement of the immune responses. In: Samter M, Talmage DW, Frank MM, Austen
IRON IN ITAI-ITAI DISEASE
12.
13. 14.
IS.
16.
17.
18.
19.
20.
21.
22.
23.
24. 25.
KF, Claman HN (eds.) Immunological Diseases, 4th ed., Vol. I . Little, Brown, Boston, pp. 361-384. Osserman EF, Lawlor D P 1966 Serum and urinary lysozyme (muramidase) in monocytic and monomyelocytic leukemia. J Exp Med 124:92 1-95 1. Shigematsu I (ed.) 1989 Itai-itai Disease and Cadmium Poisoning. Hoken Kankyo Report, Vol. 56. (Japanese). Yoshiki S, Ueno T, Akita T, Yamanouchi M 1983 Improved procedure for histological identification of osteoid matrix in decalcified bone. Stain Technol 58:85-89. Pearse AGE (ed.) 1985 Inorganic constituents and foreign substances. In: Histochemistry. Theoretical and Applied, Vol. 2. Analytical Technology, Chap. 20. Churchill Livingston, London, pp. 973-1033. Villanueva AR, Mehr LA 1977 Modification of the Goldner and Gomori one-step trichrome stain for plastic-embedded thin sections of bone. Am J Med Technol 43536-538. Noda M, Kitagawa M, Tomoda F, lida H 1990 Thromboiic thrombocytopenic purpura as a complicating factor in a case of polymyositis and Sjogren’s syndrome. Am J Cin Pathol 94:217--221. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier P J , Ott SM, Recker RR 1987 Bone histomorphometry: Standardization of nomenclature, symbols and units. J Bone Miner Res 2:595-610. Miyahara T , Yamada H, Takeuchi M, Kozuka H, Kato T, Sudo H 1988 Inhibitory effects of cadmium on in vitro calcification of a clonal osteogenic cell, MC3T3-EI. Toxicol Appl Pharmacol 9652-59. Ott SM, Maloney NA, Coburn JW, Alfrey AC, Sherrard DJ 1982 The prevalence of bone aluminum deposition in renal osteodystrophy and its relation to the response of calcitriol therapy. N Engl J Med 307:709-713. Phelps KR, Vigorita VJ, Bansal M, Einhorn TS 1988 Histochemical demonstration of iron but not aluminum in a case of dialysis-associated osteomalacia. Am J Med 84:775-780. Mukawa A 1985 Changes of heavy metal elements in the liver and kidney by cadmium exposure in autopsy cases with itaiitai disease and cadmium-polluted inhabitants. In: Shigematsu I (ed.) Hoken kankyo report, Vol. 51, pp. 196-199 (Japanese). Fairbanks VF, Beutler E 1984 Iron metabolism. In: Williams W J , Beutler E, Erslev AJ, Lichtman MA (eds.) Hematology, 3rd ed., Chap. 33. McGraw-Hill, New York, pp. 300-310. Hamilton DL, Valberg LS 1974 Relationship between cadmium and iron absorption. Am J Physiol 227:1033-1037. Christoffersen J , Christoffersen MR. Larsen R, Rostrup E,
255
26. 27.
28. 29.
30.
31.
32.
33.
34. 35.
36.
Tingsgaard P , Andersen 0, Grandjean P 1988 Interaction of cadmium ions with calcium hydroxyapatite crystals: A possible mechanism contributing to the pathogenesis of cadmiuminduced bone diseases. Calcif Tissue Int 42331-339. Selvig AK, Hake A 1975 The ultrastructural localization of iron in rat incisor enamel. Scand J Dent Res 83:88-95. Bauminger E, Ofer S, Gedalia 1, Horowitz G, Mayer I 1985 Iron uptake by teeth and bones: A Mossbauer effect study. Calcif Tissue Int 37:386-389. Aisen P 1980 Iron transport and storage proteins. Annu Rev Biochem 49:357-393. Schafer SG, Forth W 1984 Effect of acute and subchronic exposure to cadmium on the retention of iron in rats. J Nutr 114:1989- 1996. Meyer JL, Thomas WC 1982 Trace metal-citric acid complexes as inhibitors of calcification and crystal growth. I. Effects of Fe(IIl), Cr(lI1) and AI(II1) complexes on calcium phosphate crystal growth. J Urol 128:1372-1375. Laeng H, Egger T , Roethlisberger C , Cottier H 1988 Stainable bone iron in undecalcified, plastic-embedded sections. Occurrence in man related to the presence of “free” iron? Am J Pathol 131:344-350. Hays EF, Margaretten N 1985 Long-term oral cadmium produces bone marrow hypoplasia in mice. Exp Hematol 13: 229-234. Huebers HA, Huebers E, Csiba E, Rummel W , Finch CA 1987 The cadmium effect on iron absorption. Am J Clin Nutr 45:1007-1012. Dusso AS, Puche RC 1985 The effect of la.25-dihydroxycholecalciferol on iron metabolism. Blut 51: 103-108. Baker LRI, Barnett MD, Brozovic B, Cattell WR, Ackrill P , McAIister J , Nimmon C 1976 Hemosiderosis in a patient on regular hemodialysis: Treatment by desferrioxamine. Clin Nephrol 6:326-328. Chang TMS, Barre P 1983 Effect of desferrioxamine on removal of aluminum and iron by coated charcoal haemoperfusion and haemodialysis. Lancet 2: 105 1- 1053.
Address reprint requests to: Makofo Nodu, M.D. Deparlmenr of Internal Medicine St. Teresa Hospital 1-2-I , Koshigoe, Kamakuru, Kanaguwa, Japan, 248 Received for publication March 22. 1990; in revised form Sepiember 9, 1990; accepted October 5, 1990.
Iron as a Possible Aggravating Factor for Osteopathy in Itai-itai Disease, a Disease Associated with Chronic Cadmium Intoxication MAKOTO NODA, MASANORI YASUDA, and MASANOBU KITAGAWA
ABSTRACT
Ifai-itai disease is thought to be the result of chronic cadmium (Cd) intoxication. We examined 23 autopsy cases of itai-itai disease and 18 cases of sudden death as controls. Urine and blood samples from 10 patients were collected before they died and revealed the presence of severe anemia and renal tubular injuries. Undecalcified sections of iliac bone were stained with Aluminon reagent, an ammonium salt of aurintricarboxylic acid, and Prussian blue reagent in all cases of itai-itai disease. These two reagents reacted at the same mineralization fronts. X-ray microanalysis revealed the presence of iron at mineralization fronts in itai-itai disease. Five patients showed evidence of hemosiderosis in the liver, spleen, and pancreas, probably as a result of post transfusion iron overload. Renal calculi and calcified aortic walls were also stained with Prussian blue reagent in several patients. Neither ferritin nor transferrin were visualized at mineralization fronts in itai-itai disease by immunohistochemical staining. These results suggest that iron is bound to calcium or to calcium phosphate by a physicochemical reaction. A marked osteomalacia was observed in 10 cases of itaiitai disease by histomorphometry. Regression analyses of data from cases of itai-itai disease suggested that an Aluminon-positive metal inhibited mineralization and that renal tubules were injured. Since bone Cd levels were increased in itai-itai disease, it is likely that renal tubules were injured by exposure to Cd. Therefore, stainable bone iron is another possible aggravating factor for osteopathy in itai-itai disease, and a synergistic effect between iron and Cd on mineralization is proposed.
INTRODUCTION sometimes have serious deleterious effects on bone systems.") h i - i f u i ("ouch-ouch'' in English) disease (IID) is thought to be the result of chronic cadmium (Cd) intoxication, and patients suffer from severe osteomalacia or osteoporosis, or both, with high frequency.") It is well known that idiopathic hemochromatosis and 6-thalassemia are associated with osteopathy, and posttransfusion iron overload has been suggested as a cause of Aluminum (Al) was found in patients with dialysis-associated osteomalacia as a toxic agent for osteopathy. ( 5 ) Recently, Pierided6)suggested that iron can also be found in some patients with dialysis-associated osteomalacia. In a previous report we described how both osteoblast
T
RACE METALS
function and mineralization were impaired in patients with IID, as indicated by quantitative bone histology, and suggested that Cd may be one of the causative It is not generally recognized that undecalcified bone sections from patients with IID react with Aluminon reagent, an ammonium salt of aurintricarboxylic acid, at mineralization Aluminon reagent reacts primarily with tissue Al,'8.9)but this reagent also cross-reacts with tissue iron and other metals.(8~1a) In this study we showed the presence of stainable bone iron in 23 cases of IID, and we suggested that iron is a possible aggravating factor for osteopathy in IID.
MATERIALS AND METHODS
Selection of patients A total of 23 autopsy cases of IID, patients aged 62-84
Department of Pathology, Toyama Medical and Pharmaceutical University, Faculty of Medicine, Toyama, Japan.
245
NODA ET AL.
246 years (mean + SD, 76.6 f 5 . 9 , were examined: 1 man was aged 80 years, and 22 women were aged 62-84 years (76.4 f 5.5 years). The clinical diagnosis is described in our previous The causes of death were bronchopneumonia (7 cases), hypovolemic shock due to gastroduodenal ulcer (5 cases), lung cancer (1 case), pulmonary abscess (1 case), uremia (1 case), sepsis (1 case), urinary tract infection (3 cases), colitis (1 case), and asphyxia (3 cases). For the controls, 18 autopsy cases of women aged 60-91 years (75.7 f 8.3 years) were selected. Each had died suddenly, and causes of death included rupture of cerebral aneurysm (3 cases), burn (1 case), accident (4 cases), rupture of dissecting aneurysm ( 1 case), thrombosis of superior mesenteric artery ( 2 cases), and asphyxia (7 cases). None had a history of prolonged bed rest, immobilization, or any organic, endocrine, or metabolic disease. No glucocorticoids, antacids, or diuretics had been prescribed except at the terminal stage. Blood and urine samples were collected, before they died, from 10 of the 23 cases of IID. These 10 were all women aged 62-84 years (75.1 f 6.4). Serum concentrations of calcium, inorganic phosphorus, alkaline phosphatase, and creatinine and urinary concentrations of inorganic phosphorus and creatinine were measured by standard automated methods. Urinary fi,-microglobulin was measured by a single radial immunodiffusion method.'") Urinary lysozyme was measured by the lysoplate method."z' Urinary and bone levels of Cd were determined by flameless atomic absorption spectrophotometry after extraction with ammonium pyrolidine dithiocarbamate (APDC) and methyl isobutyl ketone (MIBK).'") Control values of urinary and serum chemistries were taken from a report by S h i g e m a t s ~ , (who ~ ~ ) summarized extensive epidemiologic and clinical investigations of inhabitants of Cd-polluted and unpolluted areas. Bone levels of Cd for controls were also taken from our previous report. ( ' ) All necropsy materials were taken from a standard site, approximately 2 cm posterior and inferior to the left anterior-superior iliac spine.
Histochemistry The autopsy specimens of iliac bone were fixed in 70% ethanol, embedded in methyl methacrylate, and sectioned in 5 pm thick slices with a Jung K microtome. The autopsy specimens were also fixed in 15% buffered formalin and postfixed in a 0.5% solution of cyanuric chloride that contained 1% N-methylmorpholin for 1 day. They were then decalcified with 10% ethylenediaminetetraacetic acid in distilled water, embedded in paraffin, and sectioned into slices of less than 3 pm thickness [Yoshiki's method(14)]. Undecalcified and decalcified sections of bone were stained with Prussian blue for iron(15)and with Aluminon reagent(8-Lo)for Al and/or iron. Undecalcified sections of bone were double-stained, first with Aluminon reagent and then with Prussian blue reagent, to examine the localization of the metal reaction. Undecalcified sections of bone were stained with Villanueva's solution.(I6)Decalcified sec-
tions of bone were stained with hematoxylin and eosin according to the Yoshiki method.(*4) The kidney, liver, spleen, pancreas, and calcified aortic walls of the autopsy materials were fixed in 15% buffered formalin, embedded in paraffin, sectioned in 3 pm thick slices, and stained with hematoxylin and eosin and Prussian blue reagent.
X-ray microanalysis Undecalcified, methyl methacrylate-embedded sections of bone from patients and control subjects were placed on carbon plates, rinsed with distilled water, and allowed to dry before being coated with carbon. X-ray microanalysis (XMA) combined with scanning electron microscopy (SEM) was carried out for examination of trabecular bone using a Hitachi X-650electron microscope equipped with an energy-dispersive XMA system Kevex 7000 operated at 20 kV. We used point XMA for the detection of iron in the bone by counting x-ray pulses from all elements in a given spot. We also set a window from 6.26 to 6.64 keV and obtained iron La (6.40 keV) x-ray pulses for line XMA, which counts iron-specific pulses of x-rays along a scanning line.
Immunohisrochemistry Immunohistochemical staining was applied to samples from 23 cases of IID using the avidin-biotin-peroxidase complex (ABC) m e t h ~ d . " ~Decalcified ) sections of bone were deparaffinized in xylol and pretreated with 1% hydrogen peroxide (H,O,) for 30 minutes. They were incubated with normal goat serum for polyclonal antibodies for 20 minutes. Sections were reacted with polyclonal antibodies raised in rabbits against transferrin and ferritin (Dako Corporation, USA) for 30 minutes. Next they were treated with biotinylated goat serum raised against rabbit immunoglobulin and then reacted with ABC (Vector Laboratories, Burlingame, CA) for 30 minutes. All peroxidase reactions were localized by reaction with 3,3'-diaminobenzidine tetrahydrochloride (Wako, Japan) that contained 0.001 9'0 H,O,. Sections were counterstained with methyl green. The specificity of immunoreactivity was confirmed by replacing the antibody with control rabbit serum.
Histomorphometry Eroded surface, quiescent surface, osteoblasts, and osteoclasts have been defined and described elsewhere.(Is) The entire area of cancellous tissue in the bone sections was quantitatively analyzed by a semiautomatic rnethod,l7) and at least 60 optical fields were analyzed. Histomorphometric measurements are described with reference to the nomenclature of Parfitt and colleagues,''8) as follows: ( I ) bone volume, the fraction of tissue volume occupied by trabecular bone (BV/TV); ( 2 ) osteoid volume, the fraction of tissue volume occupied by osteoid (OVITV); (3) mineralized volume, the fraction of tissue volume occupied by mineralized bone (Md.V/TV); (4) wall thickness, the mean
IRON IN ITAI-ITAI DISEASE distance between cement lines and the trabecular surfaces of completed structural units (W.Th); (5) osteoid thickness, the mean distance between mineralization fronts and osteoid surfaces (0.Th); (6) osteoblast surface, the fraction of osteoid surface covered by osteoblasts (Ob.S/OS); (7)osteoid surface, the fraction of trabecular surface covered by osteoid (OS/BS); (8)osteoclast surface, the fraction of eroded surface covered by osteoclasts (Oc.S/ES); (9)eroded surface, the fraction of trabecular surface occupied by eroded surface (ES/BS); (10) Aluminon-stainable bone, the fraction of trabecular surface covered with Aluminon (Aluminon/BS); and (1 1) Aluminon-stainable bone (osteoid referent), the fraction of osteoid surface stained with Aluminon (Aluminon/OS). The statistical significance of differences was determined by Student's t-test for the two group means of IID and control subjects. A covariance analysis with regression was applied to laboratory and histomorphometric data to determine correlation coefficients.
247 centrations of 1,25-dihydroxycholecalciferol [ 1,25(OH),D,] in this study. These results showed that cases of IID were associated with the presence of severe anemia and damage to renal tubules.
Histochemistry
Figures 1 and 2 show the results of staining with Aluminon and Prussian bue. Both reagents reacted at mineralization fronts in undecalcified sections of bone from patients with IID. These two reagents occasionally reacted on osteoid-free surfaces of trabecular bone. Since Aluminon reagent reacts with tissue Al as a reddish line and Prussian blue reagent reacts with tissue iron as a bluish line, it is possible to discriminate Al from iron if appropriate double-staining is performed on the same section. A clear, reddish line was observed in a case of IID at a mineralization front stained with Aluminon reagent (Fig. 3A), and a bluish line overlapped at the identical front after staining the same section with Prussian blue reagent (Fig. 3B). Similar results were obtained by double-staining with Aluminon and Prussian blue in undecalcified sections of bone from RESULTS 23 patients with IID. These results suggest the possibility Urinary and serum chemistry that Aluminon reagent reacts with tissue iron at mineralThe results of the analysis of urinary and serum chemis- ization fronts of undecalcified sections of bone in IID. Detries are listed in Table 1. We did not apply statistical treat- calcified sections of bone from the patients with 11D did ment to the analysis of the biochemical data from the pa- not react with Aluminon or with Prussian blue reagents. tients and control subjects. However, there were marked Decalcified and undecalcified sections of bone from the decreases in hemoglobin, hematocrit, percentage tubular control subjects did not react with either of these two rereabsorption of phosphate, and increases in urinary P2-mi- agents. Excess tissue hemosiderin was observed in the liver, croglobulin, urinary lysozyme, serum alkaline phosphatase, and serum creatinine. We observed the increase in spleen, and pancreas from five patients with IID after bone Cd content in IID, but we did not measure the net Prussian blue staining. No signs of idiopathic hemochrocontent of Al in this study. We did not measure the con- matosis were confirmed clinically or histologically. We
TABLE 1. URINARY A N D SERUM CHEMISTRY IN 10 PATIENTS WITH
Variable Hemoglobin, g/dl Hematocrit, 070 Urinary &-microglobulin, mg/dl Urinary lysozyme, mg/dl Urinary creatinine, mg/dl Urinary inorganic phosphorus, g/dl Urinary cadmium content, pg/liter Tubular reabsorption of phosphate, 070 Serum alkaline phosphatase, BLU/liter Serum calcium, mg/dl Serum inorganic phosphorus, rng/dl Serum creatinine, mg/dl Bone cadmium content, p g / g wet weight
Controlb
13.3 f 38.9 * 0.18 f 0.71 * 36.4 f 19.1 f 5.47 f 84.1 + 2.76 f 9.76 f 3.64 f 0.93 f 0.50
f
1.5 3.9 0.45 1.92 25.7 16.4 3.84 5.2 0.78 0.36 0.47 0.28 0.50
ITAI-ITAI
DISEASEa
Itai-itai disease (n
= 10)
6.5 f 0.5 20.3 f 2.7 3.63 + 1.23 6.29 + 3.67 31.2 f 7.0 16.2 f 5.0 4.98 + 2.03 27.2 f 8.4 5.06 f 2.09 8.37 f 0.96 3.74 + 1.16 4.69 f 1.55 2.24 f 0.87
aMean f SD. hControl values for the parameters of urinary and serum chemistry are taken from Shigemat~u,"~' and bone cadmium contents are taken from our previous report"' with the help of Dr. Kuzuhara from the National Institute of Public Health.
248
NODA ET AL.
I
FIG. 1. Photomicrograph of an undecalcified section of bone from a patient with itai-itai disease stained with Aluminon reagent. Note clear (reddish) lines located at the osteoid/bone interface. Aluminon stain, x 20 (original magnification).
2#
FIG. 2. Photomicrograph of an undecalcified section of bone from a patient with itai-itai disease stained with Prussian blue reagent. Note clear (bluish) lines located at the osteoid/bone interface as in Fig. 1. Prussian blue stain, x 20 (original magnification).
could not clarify the exact amount of iron loaded by therapeutic measure in this retrospective study, and thus we could not clarify dose-response relationships between the total amount of loaded iron and the extent of bone lesions. However, there is a possibility that some of the patients received red blood transfusions or parenteral iron or both
for severe anemia. Hemosiderin-laden erythroblasts in bone marrow always reacted with Prussian blue reagent, irrespective of whether bone sections were decalcified. Erythroblasts did not react with Aluminon reagent in decalcified or undecalcified sections of bone. Mild nephrocalcinosis was observed in several cases of
249
IRON IN ITAI-ITAI DISEASE
A
-
B g X‘
+ FIG. 3. Photomicrographs of an undecalcified section of bone from a patient with itai-itai disease stained with both Aluminon and Prussian blue reagents. (A) Aluminon staining only; and (B) Prussian blue staining after the Aluminon staining shown in A. Note the overlap between the reddish line (Aluminon stain) and the bluish line (Prussian blue stain) at mineralization fronts. Aluminon and Prussian blue stain, x 100 (original magnification).
IID, probably due to excessive dietary intake of vitamin D administration for treatment of severe osteoporosis. These renal calculi were stained with Prussian blue reagent (Fig. 4). Calcified aortic walls were also stained with Prussian blue reagent (Fig. 5 ) . These results suggest that iron is bound to calcium or to calcium phosphate at mineralization fronts, in renal calculi, in aortic walls, and presurnably in other calcified tissues of patients with 11D.
X-ray microanalysis Figure 6 shows a scanning electron micrograph of trabecular bone from a patient with IID. XMA revealed that undecalcified sections of bone contained unusual amounts of iron (Fig. 7). This result was reproduced in all cases of IID. Aluminum and Cd were not detected in any cases of IID at a significant level by this technique. Careful point XMA revealed that the counts due to iron were highest at mineralization fronts. These results were supported by the results of line XMA, which produced clear biphasic peaks (Fig. 6 ) . These peaks coincided with the distributions of
iron at mineralization fronts where both Prussian blue and Aluminon reagents reacted. Iron, Al, and Cd were not detected in control subjects at any significant level by XMA. These results suggest that undecalcified sections of bone from patients with IID contain iron that is deposited predominantly at mineralization fronts.
Immunohistochemistry Immunohistochernical staining revealed that ferritin-specific and transferrin-specific antibodies failed to react at mineralization fronts but reacted with hernatopoietic cells in bone marrow. These results suggest that trabecular bone in IID does not contain ferritin or transferrin at mineralization fronts. These results also suggest that iron is bound to calcium or calcium phosphate at mineralization fronts as “free” ions.
Histomorphometry Table 2 shows the results of histomorphometric data from 10 cases of IID and 18 control subjects. In the case of
250
NODA ET AL.
FIG. 4. Photomicrograph of a renal section from a patient with itai-itai disease stained with Prussian blue. Renal calculi are stained with Prussian blue reagent and visible as fine granular deposits. Prussian blue stain, x 50 (original magnification).
FIG. 5. Photomicrograph of a calcified aortic wall from a patient with itai-itai disease stained with Prussian blue stain. Calcified tissue are stained with Prussian blue reagent and visible as fine granular deposits. Prussian blue stain, x 50 (original magnification).
all parameters of bone structure and of the formation and resorption of bone, significant differences were found between group means for the patients and control subjects. Osteoid volume, osteoid thickness, osteoblast surface, 0steoid surface, and eroded surface were significantly greater in patients (p < 0.01). Mineralized volume and wall thick-
ness were significantly lower (p < 0.01). The significant increase in bone volume (p < 0.01) was due to an increase in osteoid volume. These results suggest that patients with IID experience a marked osteomalacia with or without reduction in bone mass. The results of regression analyses of data from 10 cases
25 1
IRON IN ITAI-ITAI DISEASE
FIG. 6. Scanning electron micrograph of trabecular bone with excess osteoid accumulation from a patient with itaiitai disease. The upper straight line is a scanning line. The lower zigzag line indicates the result of line XMA, which counts iron-specific pulses of x-rays along a scanning line. Two peaks (arrows) are seen at the mineralization fronts of the trabecular bone. Asterisk (*) indicates the spot at which point XMA was carried out. x 1OOO.
that the area of Aluminon-stained surface with osteoid referent was inversely correlated with osteoid volume ( r = -0.504) and osteoblast surface ( r = -0.490). The bone Cd content was inversely correlated with urinary inorganic phosphorus ( r = -0.624, p < 0.05). Table 4 shows a list of coefficients of correlation between histomorphometric data and laboratory data. Urinary lysozyme increases with osteoid volume ( r = 0.738, p < 0.05), and serum inorganic phosphorus also increases with both osteoid volume ( r = 0.767, p < 0.01) and osteoid thickness ( r = 0.683, p < 0.05). These results suggest that if renal tubules are injured and, thus, 1,25-(OH),D, is not produced adequately because of injury to renal tubules, reabsorption of lysozyme, excretion of inorganic phosphorus, and bone mineralization are impaired. If the percentage tubular reabsorption of phosphate decreases, as is frequently found in IID, the osteoblast surface increases ( r = -0.656, p < 0.05). Since Cd usually inhibits osteoblast f u n ~ t i o n , ~ ~this ~ ' ' 'result suggests the possibility that Cd, if present, may affect renal tubules directly. Osteoclast surface was correlated positively both with serum creatinine ( r = 0.794, p < 0.01) and with serum alkaline phosphatase ( r = 0.488), suggesting that the increase in osteoclast surface may relate to the advances in renal insufficiency. Osteoclast surface was also correlated positively with Aluminon-stained surface, as shown in Table 3 ( r = 0.651, p < 0.05). Aluminon-positive metal may affect the progress of renal insufficiency or bone resorption or both, although we did not measure concentrations of serum parathyroid hormone in this study. The present histomorphometric data indicate that iron may impair bone mineralization and osteoblast function in cases of IID with osteomalacia. Furthermore, Cd may injure renal tubules directly.
DISCUSSION
This study showed the presence of iron at mineralization fronts in undecalcified sections of bone from patients with IID. This result was clearly demonstrated by staining with Prussian blue, which reacts with tissue iron with a high deof IID are shown in Tables 3 and 4. Table 3 shows a list of gree of sensitivity and sepcificity,"') and confirmed by correlation coefficients between Aluminon- or Cd-related point and line XMA. Undecalcified sections of bone also data and histomorphometric and laboratory data. Alumi- reacted with Aluminon reagent in all cases with 1ID. Alunon-stained surface was correlated inversely with mineral- minon reagent reacts primarily with Al, and Al is thought ized volume ( r = -0.743, p < 0.05) and positively with to be a causative metal for osteomalacia associated with osteoid surface ( r = 0.740, p < 0.05), suggesting that Alu- dialysis." ' Aluminon reagent also cross-reacts with tisminon-positive metal inhibited mineralization. This sug- sue iron.'8 l o ) We observed that both Aluminon and Prusgestion is supported by the result that Aluminon-stained sian blue reagents reacted at the same mineralization fronts surface area was correlated inversely with wall thickness ( r by double-staining of undecalcified section of bone. We = -0.501) and bone volume (r = -0.495). Aluminonalso observed that bone Cd content in IID was increased in stained surface with osteoid referent defines the fraction of this study and in our previous report,'7) although Cd and osteoid surface that is covered with Aluminon, and this pa- Al were not detected by XMA at significant levels in this rameter was correlated inversely with osteoid thickness (r study. It therefore appears that bones of patients with IID = -0.682, p < 0.05). Thus if the area of Aluminon- contain iron as well as Cd, and that iron, like Al in dialysis stained surface increases, then osteoid thickness decreases osteomalacia, may play important roles in osteopathy. within an estimated osteoid area. This result suggests that Idiopathic heniochromatosis rarely complicates osteoAluminon-positive metal inhibits the formation of osteoid penia.I3) De Vernejoul et al.'" reported a hyperosteoidosis by osteoblasts. This suggestion is supported by the result and mineralization defect in patients with 0-thalassemia.
252
NODA ET AL.
FIG. 7. Spectra of point XMA of the trabecular bone seen in Fig. 6 at the spot marked *. The spectrum in A shows a wide range of energies, and the spectrum in B shows a narrow range of the energies shown in A. Note the presence of iron and the absence of Cd and Al.
TABLE 2. HISTOMORPHOMETRIC DATAFOR 10 PATIENTS WITH ITAI-ITAIDISEASE^ ~~
Variable
Control(n
Bone volume, To Osteoid volume, Vo Mineralized bone volume, 070 Wall thickness, pm Osteoid thickness, pm Osteoblast surface, 070 Osteoid surface, Vo Osteoclast surface, Vo Eroded surface, Vo Aluminon-stainable surface, Vo Aluminon-stainable surface, To (osteoid referent)
10.3 0.10 10.2 51.4 10.0 3.10 10.1 8.40 8.31
=
18)
f 4.8 f
f f f
= f
+ f
0.08 4.8 7.2 2.1 4.49 7.4 9.10 7.28
Not stained Not stained
(n
=
IO)
f 6.8b 5.92 + 5.19b 5.8 f 5.Ob 46.1 f 4.9b 41.3 + 19.8b 3.84 f 3.37b 83.4 f 9.8b 23.07 + 13.39b 9.04 f 3.93b 43.1 + 10.0 55.9 f 12.2
11.7
aMean f SD. bp < 0.01.
Recently Pieridesc6)reported the association between iron overload and osteomalacia in hemodialysis patients. Phelps et al.(”] found tissue iron but not A1 in a case of dialysis osteomalacia. These authors showed iron at mineralization fronts by Prussian blue staining and suggested that the iron was brought about by overdose of dialysate or red
blood transfusions. Severe anemia was present in our patients. Five patients also showed evidence of posttransfusion hemosiderosis in the liver, spleen, and pancreas. These patients may have received an overdose of exogenous iron supplements. Many patients suffer from severe anemia due to various
253
IRON IN ITAI-ITAI DISEASE
TABLE 3. CORRELATION COEFFICIENTS BETWEENLABORATORY AND HISTOMORPHOMETRIC DATAAND ALUMINONAND CADMIUM-RELATED DATAIN 10 CASESOF ITAI-ITAIDISEASE
Aluminon-stainable surface, Yo Aluminon-stainable surface, Yo (osteoid referent) Urinary Cd content
Bone Cd content
ar,
Wall thickness, -0.501 Bone volume, -0.495
Mineralized volume, -0.743b Osteoid surface, 0.740b Osteoclast surface, 0.651b Osteoid thickness, -0.682b Bone volume, -0.606b
Urinary creatinine, -0.705h Urinary inorganic phosphorus, -0.624b
Osteoid volume, -0.504 Osteoblast surface, -0.490 Serum calcium, 0.402 Osteoid volume, -0.533 Urinary lysozyme, -0.496 Wall thickness, -0.460 Serum inorganic phosphorus, -0.457 Osteoclast surface, -0.525
coefficient of correlation.
bp < 0.05.
TABLE4. CORRELATION COEFFICIENTS BETWEEN HISTOMORPHOMETRIC DATAAND LABORATORY DATA IN 10 CASESOF ITAI-ITAI DISEASE
Urinary &-microglobulin Urinary lysozyme
Osteoid volume, 0.738b
Urinary creatinine Urinary inorganic phosphorus To Tubular reabsorption of phosphate
Eroded surface, -0.636b Osteoblast surface, -0.656h
Serum alkaline phosphatase Serum inorganic phosphorus
Eroded surface, 0.863~ Osteoid volume, 0.767~ Osteoid thickness, 0.683b Osteoid thickness, -0.614b Osteoclast surface, 0.794" Osteoblast surface, 0.655b
Serum calcium Serum creatinine
ar,
Osteoblast surface, 0.437 Eroded surface, -0.419 Bone volume, 0.549 Osteoid thickness, 0.536 Wall thickness, 0.47 1 Osteoclast surface, 0.441 Mineralized volume, -0.575 Osteoid thickness, -0.486 Osteoclast surface, 0.488 Osteoid surface, 0.493 Osteoclast surface, 0.432 Osteoid volume, -0.556 Eroded surface, 0.489 Osteoid thickness, 0.472 Mineralized volume, -0.441
coefficient of correlation.
bp c: 0.05. ' p < 0.01.
underlying diseases and receive repeated red blood transfusions and/or parenteral iron, but the frequency of iron-induced osteopathy is very low. There were no patients in our study who had undergone hemodialysis. Some patients with IID had severe osteomalacia and osteoporosis, but they did not reveal clinical signs of anemia or past history of red blood transfusions. ' z ) Concentrations of iron were reported to be normal or even low in the liver, pancreas, and kidney of people living in Cd-polluted areas.'22)
Therefore, these observations d o not necessarily point to iron as a single etiologic agent for osteopathy in IID. Body iron is usually bound by or incorporated into various proteins, for example hemoglobin iron, storage iron, and transport iron, and it does not exist as a free ati ion."^' If iron-deficiency anemia is caused by malnutrition, multiple pregnancies, and other factors, as in IID, and if exogenous iron is not supplied, it is unlikely that body iron shifts to bone as a specific storage site. How-
NODA ET AL.
254
ever, in this study stainable bone iron was clearly observed in all cases examined. The reason for stainable bone iron in iron-deficient patients with IID is currently unknown, but it may be explained in part by the damage to tissue caused by Cd. Itai-itai disease was observed in inhabitants of Cdpolluted areas, and we showed an increase in bone Cd content in patients with IID. There is a possibility that patients in this study were exposed to Cd. The damage to renal tubules was demonstrated by the results of urinary and serum chemistries and of histomorphometry. Since serum chemistries showed advanced renal insufficiency in IID, the decrease in the percentage tubular reabsorption of phosphate alone may explain the increased phosphaturia per nephron and does not necessarily indicate the presence of injury t o renal tubules. We did not measure phosphorus excretion in patients exposed to Cd before they develop chronic renal insufficiency. However, an injury to renal tubules is a frequent finding in clinical and experimental studies of Cd intoxication.I2’ Therefore, renal tubules in IID seemed t o be injured by exposure to Cd, and tissue damage caused by Cd may relate to the onset and progress of iron accumulation in bone. Cadmium appears to compete with iron at the transfer system in the intestinal wall, and absorption of Cd increases when iron-deficiency anemia is present.(z4’Christoffersen et aI.(”) reported that Cd ions were adsorbed onto crystals of calcium hydroxyapatite and incorporated into the crystals, making them very resistant to subsequent dissolution. Some investigators have reported that iron ions can be incorporated in hydroxyapatite crystals or absorbed on the surface of teeth and bone^.^^^^"^ These reports suggest that both C d and iron ions affect the growth of crystals of calcium hydroxyapatite and contribute to the pathologic changes in the bone tissue. Both Cd and iron belong to a group of multivalent heavy metals and are bound to plasma transferrin with the same stoichiometry.‘z8.29) Meyer et al.‘Jo)reported that there is a synergistic effect between trivalent iron and citric acid on the growth of crystals of calcium phosphate, and that an iron-citrate complex may inhibit biologic calcification. Therefore, the synergistic effect not only of iron but also of Cd on the growth of crystals of calcium hydroxyapatite is more likely to impair bone mineralization than the effect of each metal alone. The results from histochemistry and immunohistochemistry suggest that iron is present as “free” ions and is bound to calcium or calcium phosphate as the result of a physicochemical reaction. Recently, Laeng et a1.(31)postulated the presence of an osteoid pool of iron in patients with primary hemochromatosis as an unusual site of iron storage. It is well known that long-term administration of Cd produces bone marrow hypoplasia and iron deficiency and thus decreases tissue utilization of Heubers et al.1331reported the effect of Cd on the release of iron from transferrin in iron-deficient rats. Therefore it is possible that body iron could shift to the bone as the result of a physicochemical reaction and that bone would then serve as an unusual site for storage of iron, when tissue utilization of iron is altered by pre-existing exposure to Cd and/ or other toxic metals. If this hypothesis is valid, long-term
therapy with iron supplements and blood transfusions for severe anemia and administration of 1 ,25-(OH),D3, which induces iron uptake,1341as a treatment for osteoporosis may modify the clinical course of IID. Alternatively, desferrioxamine may be of great value for the removal of iron and other metals in this d i s e a ~ e . ‘ ~ ~ . ~ ~ ) In conclusion, stainable bone iron was clearly demonstrated by histochemistry and XMA, and injuries to renal tubules were suggested in all cases with 11D. The stainable bone iron is a possible aggravating factor for osteopathy in IID, and a synergistic effect between iron and Cd on bone mineralization seems likely.
ACKNOWLEDGMENTS This study was performed in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Ph.D.) at Toyama Medical and Pharmaceutical University. We thank Drs. A. Miwa and M. Murai and Mr. H . Yamane for their efforts in subject sampling and their encouragement. We are grateful to Dr. Y. Kuzuhara of the National Institute of Public Health for contributing the measurements of the levels of Cd in bone. We express our thanks to Mr. T. Kumada and Mr. M. Kawahara for excellent technical assistance. We also thank Mr. T. Nagata and Mr. E. Tahara for preparation of tissues.
REFERENCES 1. Priest ND (ed.) 1985 Metals in Bone. M T P Press Limited,
Lancaster. 2. Nogawa K 1981 Itai-itai disease and follow-up studies. In: Nriagu J O (ed.) Cadmium in the Environment. John Wiley & Sons, New York, pp. 1-37. 3. Delbarre F 1960 L’Osteoporose de I’hemochromatose. Semin Hop Paris 36:3279-3284. 4. De Vernejoul MC, Girot R, Gueris J , Cancela L, Bang S, Bielakoff J , Mautalen C, Goldberg D, Miravet L 1982 Cal-
cium phosphate metabolism and bone disease in patients with hornozygous thalassemia. J Clin Endocrinol Metab 54:276281. 5 . Ellis HA, McCarthy J H , Herrington J 1979 Bone aluminum
6. 7.
8. 9.
in haernodialysed patients and in rats injected with aluminum chloride: Relationship to impaired mineralization. J Clin Pathol 32:832-839. Pierides AM 1984 Iron and aluminum osteomalacia in hemodialysis patients (letter). N Engl J Med 310:323. Noda M, Kitagawa M 1990 A quantitative study of iliac bone histopathology on 62 cases with itai-itai disease. Calcif Tissue Int 47:M-74. Hammett LP, Sottery C T 1925 A new reagent for aluminum. J Am Chem SOC47:142-143. Maloney NA, Ott SM, Alfrey AC, Miller NL, Coburn JW, Sherrard DJ 1982 Histological quantitation of aluminum in iliac bone from patients with renal failure. J Lab Clin Med 99:206-216.
10. Savory J, Wills MR 1986 Analytical methods for aluminum measurement. Kidney Int 29524-27. 11. Crowle AJ 1988 Detection and measurement of the immune responses. In: Samter M, Talmage DW, Frank MM, Austen
IRON IN ITAI-ITAI DISEASE
12.
13. 14.
IS.
16.
17.
18.
19.
20.
21.
22.
23.
24. 25.
KF, Claman HN (eds.) Immunological Diseases, 4th ed., Vol. I . Little, Brown, Boston, pp. 361-384. Osserman EF, Lawlor D P 1966 Serum and urinary lysozyme (muramidase) in monocytic and monomyelocytic leukemia. J Exp Med 124:92 1-95 1. Shigematsu I (ed.) 1989 Itai-itai Disease and Cadmium Poisoning. Hoken Kankyo Report, Vol. 56. (Japanese). Yoshiki S, Ueno T, Akita T, Yamanouchi M 1983 Improved procedure for histological identification of osteoid matrix in decalcified bone. Stain Technol 58:85-89. Pearse AGE (ed.) 1985 Inorganic constituents and foreign substances. In: Histochemistry. Theoretical and Applied, Vol. 2. Analytical Technology, Chap. 20. Churchill Livingston, London, pp. 973-1033. Villanueva AR, Mehr LA 1977 Modification of the Goldner and Gomori one-step trichrome stain for plastic-embedded thin sections of bone. Am J Med Technol 43536-538. Noda M, Kitagawa M, Tomoda F, lida H 1990 Thromboiic thrombocytopenic purpura as a complicating factor in a case of polymyositis and Sjogren’s syndrome. Am J Cin Pathol 94:217--221. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier P J , Ott SM, Recker RR 1987 Bone histomorphometry: Standardization of nomenclature, symbols and units. J Bone Miner Res 2:595-610. Miyahara T , Yamada H, Takeuchi M, Kozuka H, Kato T, Sudo H 1988 Inhibitory effects of cadmium on in vitro calcification of a clonal osteogenic cell, MC3T3-EI. Toxicol Appl Pharmacol 9652-59. Ott SM, Maloney NA, Coburn JW, Alfrey AC, Sherrard DJ 1982 The prevalence of bone aluminum deposition in renal osteodystrophy and its relation to the response of calcitriol therapy. N Engl J Med 307:709-713. Phelps KR, Vigorita VJ, Bansal M, Einhorn TS 1988 Histochemical demonstration of iron but not aluminum in a case of dialysis-associated osteomalacia. Am J Med 84:775-780. Mukawa A 1985 Changes of heavy metal elements in the liver and kidney by cadmium exposure in autopsy cases with itaiitai disease and cadmium-polluted inhabitants. In: Shigematsu I (ed.) Hoken kankyo report, Vol. 51, pp. 196-199 (Japanese). Fairbanks VF, Beutler E 1984 Iron metabolism. In: Williams W J , Beutler E, Erslev AJ, Lichtman MA (eds.) Hematology, 3rd ed., Chap. 33. McGraw-Hill, New York, pp. 300-310. Hamilton DL, Valberg LS 1974 Relationship between cadmium and iron absorption. Am J Physiol 227:1033-1037. Christoffersen J , Christoffersen MR. Larsen R, Rostrup E,
255
26. 27.
28. 29.
30.
31.
32.
33.
34. 35.
36.
Tingsgaard P , Andersen 0, Grandjean P 1988 Interaction of cadmium ions with calcium hydroxyapatite crystals: A possible mechanism contributing to the pathogenesis of cadmiuminduced bone diseases. Calcif Tissue Int 42331-339. Selvig AK, Hake A 1975 The ultrastructural localization of iron in rat incisor enamel. Scand J Dent Res 83:88-95. Bauminger E, Ofer S, Gedalia 1, Horowitz G, Mayer I 1985 Iron uptake by teeth and bones: A Mossbauer effect study. Calcif Tissue Int 37:386-389. Aisen P 1980 Iron transport and storage proteins. Annu Rev Biochem 49:357-393. Schafer SG, Forth W 1984 Effect of acute and subchronic exposure to cadmium on the retention of iron in rats. J Nutr 114:1989- 1996. Meyer JL, Thomas WC 1982 Trace metal-citric acid complexes as inhibitors of calcification and crystal growth. I. Effects of Fe(IIl), Cr(lI1) and AI(II1) complexes on calcium phosphate crystal growth. J Urol 128:1372-1375. Laeng H, Egger T , Roethlisberger C , Cottier H 1988 Stainable bone iron in undecalcified, plastic-embedded sections. Occurrence in man related to the presence of “free” iron? Am J Pathol 131:344-350. Hays EF, Margaretten N 1985 Long-term oral cadmium produces bone marrow hypoplasia in mice. Exp Hematol 13: 229-234. Huebers HA, Huebers E, Csiba E, Rummel W , Finch CA 1987 The cadmium effect on iron absorption. Am J Clin Nutr 45:1007-1012. Dusso AS, Puche RC 1985 The effect of la.25-dihydroxycholecalciferol on iron metabolism. Blut 51: 103-108. Baker LRI, Barnett MD, Brozovic B, Cattell WR, Ackrill P , McAIister J , Nimmon C 1976 Hemosiderosis in a patient on regular hemodialysis: Treatment by desferrioxamine. Clin Nephrol 6:326-328. Chang TMS, Barre P 1983 Effect of desferrioxamine on removal of aluminum and iron by coated charcoal haemoperfusion and haemodialysis. Lancet 2: 105 1- 1053.
Address reprint requests to: Makofo Nodu, M.D. Deparlmenr of Internal Medicine St. Teresa Hospital 1-2-I , Koshigoe, Kamakuru, Kanaguwa, Japan, 248 Received for publication March 22. 1990; in revised form Sepiember 9, 1990; accepted October 5, 1990.
