- Email: [email protected]
FAILURE OF SERUM FERRITIN LEVELS TO PREDICT BONE-MARROW IRON CONTENT AFTER INTRAVENOUS IRON-DEXTRAN THERAPY
652 FAILURE OF SERUM FERRITIN LEVELS TO PREDICT BONE-MARROW IRON CONTENT AFTER INTRAVENOUS IRON-DEXTRAN THERAPY*
MAJID ALI A. OLUSEGUN FAYEMI JOSEPH FRASCINO
ROBERT RIGOLOSI EVALYNNE V. BRAUN ROBERTO SINGER
Departments of Pathology of College of Physicians and Surgeons of Columbia University, The Mount Sinai School of Medicine, New York, and Holy Name Hospital, and the Regional Hemodialysis Center, Holy Name Hospital, Teaneck, New Jersey, U.S.A. The relation between serum ferritin levels and stainable-iron deposits in the liver, bone marrow was investigated in 36 patients with and spleen, chronic renal failure who died after being on haemodialysis for 1-103 months. Elemental iron (mean, 5450 mg) had been given intravenously as iron dextran to patients in a long-term subgroup, who had been on dialysis for more than 3 months. The results of semiquantitative histochemical assessment of tissue iron in slides obtained at necropsy (scale 0 to 4+) were confirmed by chemical analyses of tissue iron. Serum ferritin levels correlated well with the degree of hepatosplenic siderosis but did not always correlate with bone-marrow iron stores in these patients. Serum ferritin concentrations were raised in 10 marrow-iron-depleted subjects (mean, 1336
Summary
ng/dl). The paradoxical association of hepatosplenic siderosis with iron depletion was observed in most of the patients on dialysis for less than 40 months. The histochemical data show that the bulk of intravenously injected iron dextran is taken up by the liver and spleen; that the hepatosplenic stores fail to be mobilised to the bone marrow; and that intravenous iron-dextran therapy, by-passing the intestinal mechanism for the regulation of iron absorption, carries a high risk of marrow
long-term hepatosplenic siderosis. Introduction
THE assay of serum ferritin concentrations to determine the requirements of haemodialysis patients has gained widespread acceptancel-4 on the basis that there is a iron
correlation between serum ferritin levels and bone-marrow iron stores. 1-3 Serum ferritin concentrations generally reflect total-body iron stores: low serum ferritin levels indicate iron deficiency,S,6 and the levels in disease states associated with tissue iron overload are high, sometimes over 5000 ng/dl. 6—8 The paradoxical coexistence of hepatosplenic siderosis and bone-marrow iron depletion has been observed in some patients with haemodialysis haemosiderosis9 and cirrhosis10, 11and in a single case each ofhaemochromatosis, 12 congenital atransferrinaemia,13 and anorexia nervosa.14 There have been reports of iron-depleted bone marrow in the presence of massive iron overload in other tissues in patients with haemophilials and pulmonary haemosiderosis.16 In patients with hepatosplenic siderosis, serum ferritin levels could theoretically be raised because of the elevated hepatosplenic iron stores even though the bone marrow is depleted of iron. This possibility has not yet been investigated with histochemical and biochemical studies. We now report data on serum ferritin and reticuloendothelial iron stores, marrow and hepatosplenic, in 36 the 13th annual meeting of the American Washington D.C., in 1980.
*Presented in part
Nephrology,
at
Society of
chronic renal failure who died after 1 to 103 months on maintenance haemodialysis. We have examined the relation between the duration of haemodialysis and the tissue iron stores in these and 19 other dialysis patients in order to clarify how intravenously administered iron dextran is used.
patients with
Patients and Methods We selected for the study 36 haemodialysis patients for whom data ferritin levels were available and whose necropsies were carried out consecutively at the regional haemodialysis centre at Holy Name Hospital, Teaneck, New Jersey. The results of an additional 19 consecutive necropsies on haemodialysis patients were included in an examination of the relation between tissue iron levels and the duration of dialysis. Clinical information including the age and sex of patients, duration of haemodialysis, quantity of intravenous iron dextran given, volume of packed cells and whole blood transfused, and other haematological data, was obtained from patients’ records. The raised serum ferritin levels had been established in most of the patients by two or more determinations separated by intervals of 1 month carried out when the patients were not receiving any blood transfusions. With few exceptions, the patients in the study had not received any iron dextran in the 6 months before death. 7 of the 36 patients had been on maintenance haemodialysis for less than 3 months and served as a control group. on serum
Serum Ferritin Serum ferritin
assayed with a double-antibody precipitation radioimmunoassay (Calbiochem-Behring kit). The mean serum ferritin levels measured with this technique in our laboratory for 20 normal volunteers and non-dialysis patients without haemotological disorders were for men, 186 ng/dl (range, 34-371 ng/dl) and for women, 117 ng/dl (range, 17-265 ng/dl). was
Morphological Studies We re-examined the routine histological sections which had been taken at necropsy, fixed with 4% buffered formaldehyde, and stained with haematoxylin and eosin. Two to four slides each of the liver, spleen, and bone marrow were available. Slides from all patients were stained with histochemical stains which use the prussian-blue reaction to detect iron. Known positive and negative control sections for iron preparations were stained with each batch of study slides. Sections of bone marrow from 3 non-dialysis patients who died without any known renal disease or disorder of iron metabolism were used for comparison when evaluating the bone-marrow sections of patients in the dialysis study group. Trichrome and reticulin stains of liver sections were studied to evaluate the reticular framework in 10 patients with advanced siderosis.
Tissue Iron Scores The iron stores in the marrow, liver, and spleen were evaluated according to a previously described9 semiquantitative system of scoring on a scale’of0 to 4 + (fig. 1). Diffuse blue stippling and fine blue granules were scored as 1 +; coarse heavy granules, sometimes coalescing to form large massive granules, were scored as 4+; and the blue granularity ranging between these two extremes was divided into moderate deposits scored as 2 + and heavier deposits scored as 3 +. Sections with no staining or rare minute foci of blue granularity were classified as negative. The marrow sections from the 3 non-dialysis patients were assigned a value of 2+ for comparison with marrow sections from dialysis patients. Iron stores found in areas of recent or old haemorrhage in these organs were excluded from this evaluation.
Quantitative Chemical Analysis of Tissue Iron All tissues for histology were fixed with neutral buffered 4% formaldehyde. The bone specimens were decalcified and fixed with a formic acid/formaldehyde solution for 24 h. To ensure that iron
653
of liver from 4 haemodialysis patients stained with prussian blue. (A). No stainable iron in liver of a patient on haemodialysis for 1 month only. (B) Iron deposits of score 1 +. (C). Iron stores representing 3+ score. (D). Massive iron deposits (4+) from a patient with advanced hepatic siderosis. Original magnification fields A and D, x40, fields B and C, x 100; all 4 are reduced by li’3.
Fig. I-Photomicrographs of sections
not dissolved from the tissue during fixation and decalcification, quantitative chemical analyses for iron were carried out on fresh and fixed tissues. For chemical analysis the air-dried tissue sample was weighed then digested with concentrated nitric acid, and the iron content in the sample was analysed by atomic absorption in a heated graphite furnace. The National Bureau of Standards’ standard no. 1577 (bovine liver, iron content 18 jug/g airdried weight) was used as the reference sample. was
Fig. 2-Correlation of semiquantitative scores haemodialysis patients. r=0’689: p
Results Clinical and Laboratory Data The 23 men and 13 women in the study died at age 46 to 76 years (mean age, 61’5years). The duration of maintenance haemodialysis ranged from 1 to 103 months; the mean for 29 patients in the long-term subgroup (>3 months) was 39 months. Negative-pressure dialysers of hollow-fibre and parallel-plate types were used for all patients from 1977 to 1979; some patients who had been on dialysis for more than 3 years had used coil-type dialysers during the preceding years. The patients in the long-term group received a mean of 9 -8 units of packed cells or whole blood during the period of treatment; blood was transfused either for control of a bleeding diathesis or when improvement in red-cell mass was clinically deemed urgent. All patients were given 100-105 mg/day iron orally as ferrous sulphate for the duration of the treatment. Patients on dialysis for more than 3 months received iron as iron dextran- (’Imferon’) intravenously (mean 5450 mg during the treatment period; range, 30014 850 mg). All 36 patients in the study were anaemic; the haemoglobin levels ranged from 5 -8 to 10 -4 g/dl (mean, 7’ 9 g/dl). Of 10 patients with marrow iron depletion in the presence of hepatosplenic siderosis, anaemia was microcytic in 8 patients and normocytic in 2; niild to moderate poikilocytosis was observed in all patients. ’
Hepatosplenic Siderosis Advanced hepatosplenic siderosis (3 + to 4 + iron deposits) was observed in 23 of the 36 patients (fig. 2). In the control group (<3 months on haemodialysis) there was no stainable iron in 3 patients and focal bluish stippling of the cytoplasm of the hepatocytes in 3 patients. In the liver iron deposits appeared first mainly in the Kupffer cells; in advanced siderosis there were coarse blue granules, at times coalescing to form large masses, within both the reticuloendothelial and
serum
for
ferritin
hepatosplenic
concentrations iron
deposits
with in 36
the hepatic parenchymal cells. In siderosis of the spleen, as in the liver, cells lining the sinusoids had the heaviest iron
deposits; malpighian corpuscles were generally spared even in advanced iron overload. Other notable histological features in the liver were mild fatty changes, central venous congestion, and an increase in the fibroconnective framework; the last was most prominent in advanced hepatic siderosis. Bone-marrow Iron Scores The distribution of marrow iron scores is shown in fig. 3. The marrow was iron-depleted (scored as negative) in 10 patients with 3+ to 4+ hepatosplenic siderosis and high serum ferritin concentrations (537-3994 ng/dl) (r=0-42;
p<0-01). Chemical Analysis
of Tissue Iron Semiquantitative histochemical estimation of tissue iron was validated by chemical analysis of iron in the liver, spleen, and bone marrow (see table). Fixed and decalcified samples of bone marrow showed larger quantities of iron than fresh samples (Sample 1: fresh bone 425
washed
However, bone-marrow iron does not appear to be any significant degree by this process.
out to
Serum Ferritin Levels ferritin concentrations correlated well with the semiquantitative iron scores for hepatosplenic iron deposits
The
serum
654
and bone-marrow iron deposits after intravenous iron-dextran therapy in 55 haemodialysis patients. Numbers in parentheses are the numbers of patients in each 10-month
Fig. 4-Hepatosplenic therapy period.
(fig. 2; r
diminished marrow-iron stores in the first 30 months of maintenance dialysis; in several patients the bone marrow was depleted. The 7 patients on dialysis less than 3 months had less iron in their bone-marrow sections than the control patients. During the first few months of dialysis the dichotomy between the hepatosplenic and the bone-marrow iron scores progressed until the paradox of hepatosplenic siderosis coexisting with marrow-iron depletion was reached in 21 patients. In many patients on dialysis for more than 30 months there were progressive increases in the stainable-iron content of the bone marrow, presumably after liver and spleen were saturated with iron. The 5 patients on dialysis for more than 80 months had massive iron deposits in the bone
did
marrow.
Fig. 3-Distribution of bone-marrow iron scores. = 0’ 689, p<0’ 001). By contrast, serum ferritin levels always correlate with the stainable iron scores in the marrow (fig. 3; r=0’42, p<0’01). Raised serum ferritin
not
found in - 10 marrow-iron-depleted 537-3994 ng/dl; mean 1336 ng/dl), in 6 of whom there was no stainable iron and in 4 there were rare minute deposits of iron. For 3 other patients with raised serum ferritin concentrations (1490-2510 ng/dl) marrow iron was scored as 2+ (the value for non-dialysis control marrow sections). In general, serum ferritin levels were higher in patients who had been on haemodialysis therapy for longer periods of time; mean values for 29 patients treated for more than 3 months and 7 patients treated for less than 3 months were 1820 ng/dl and 338 ng/dl, respectively. concentrations
were
patients (range,
Tissue Iron and Duration
of Dialysis All patients on dialysis for more than 20 months had moderate to heavy iron deposits in liver and spleen sections; there was massive siderosis of these organs in all patients on dialysis for more than 40 months. Though semiquantitative, the hepatosplenic iron scores were directly related to the duration of therapy (fig. 4). In contrast, the patients showed CORRELATION BETWEEN HISTOCHEMICAL ESTIMATION OF TISSUE IRON AND QUANTITATIVE CHEMICAL ANALYSIS IN 3 PATIENTS WITH TISSUE IRON OVERLOAD
Histochemical tissue iron
weight.
scale, 0-4+. Chemical analysis results in pg/g dry
Discussion Ferritin is the principal iron-containing protein in the tissues, and with its insoluble derivative, haemosiderin, it accounts for most of the iron identified in histological sections with the prussian-blue reaction. The ferritin molecule is thought to consist of a crystalline ferric-oxidephosphate core and a protein shell with a molecular weight of 450 000.’ Although many tissues can synthesise ferritin, the liver, spleen, and bone marrow are the primary sites of production. 17,11 Since an immunoradiometric procedure for its measurement in serum has been available 19 serum ferritin has been superior to the serum iron and serum iron-bindingcapacity tests as an indicator of marrow-iron reserves.l,3,. Serum ferritin concentrations below the normal range indicate iron deficiency;5,6,2o raised levels have been observed in disease states associated with increased body iron stores.6-8 Serum ferritin levels are raised in acute and chronic liver diseases and in some malignant disorders,21,22 as well as in disorders of iron metabolism. Our data agree with studies correlating serum ferritin concentrations and total-body iron stores in non-uraemic patients 17-20 but do not fully accord with some studies showing correlation between the bone-marrow iron and serum ferritin in haemodialysis patients. 1-3 In our study, when serum ferritin was low, the marrow was iron-depleted. but the converse did not always obtain. High serum ferritin levels in our patients with hepatosplenic siderosis were not correlated with the bone-marrow iron stores (fig. 3). In 10 patients, the serum-ferritin concentrations were high and yet there was little or no stainable iron in the bone marrow. In another 3 patients with high serum ferritin levels
655
(1490-2510 ng/dl) the marrow iron score was only 2 + (nondialysis control value). 1-3 In previous studies of serum ferritin and marrow iron, no attempt was made to investigate iron overload in the liver and spleen. In the absence ofhepatosplenic siderosis the marrowiron stores can be expected to represent total-body iron stores, and iron-depleted marrow can be regarded as evidence of iron deficiency. This is not true of dialysis haemosiderosis, where there may be a paradox of hepatosplenic iron overload and marrow iron depletion. In this setting, high serum ferritin levels are caused by hepatosplenic siderosis and may not be relied upon as indicators of marrow iron reserves. All our patients except those in the control group were given relatively large doses of iron dextran, whereas none of the patients in the study of Bell et al. received parenteral iron. Another possible cause of the different results in our study and that of Bell et awl. is that the tissue obtained by needle biopsy of bone is usually scanty and crushed and does not allow as accurate an estimate of marrow iron as do large sections of properly prepared bone obtained at necropsy. The solution used to decalcify the bone-marrow sections may cause underestimation of marrow iron, but identical histochemical procedures were used for the dialysis patients’ and controls’ sections and the results of chemical analyses of tissue iron agreed with the histochemical scores (table). The fate ofhepatosplenic iron stores after intravenous irondextran therapy has not been fully investigated. In rabbits the bulk of intravenously injected iron dextran is taken up by liver and spleen;23,24 when the animals were killed 6 months after iron-dextran overload, even larger quantities of iron were found in the liver. 24 After intravenous injection of radiolabelled iron dextran in man, more radioactivity was found in the liver than in any other organ at the end of the study period (4 to 6 weeks).25,26 Our study shows that there is a failure of mobilisation of hepatosplenic iron to the bone marrow in this setting and that hepatosplenic iron overload may persist even in the face of bone-marrow iron depletion. Our data raise serious questions about the appropriateness of intravenous iron-dextran therapy for treatment of iron deficiency in dialysis patients. They show that marrow iron cannot be replenished by intravenous iron dextran without incurring serious hepatic iron overload and the risk of liver
injury.
REFERENCES
1. Hussein S, Prieto J, O’Shea M, Hoffbrand AV, Baillod RA, Moorhead JF. Serum ferritin assay and iron status in chronic renal failure and haemodialysis. Br Med J 1975,i: 546-48. 2 Mirahmadi SM, Paul WL, Winer RL, et al Serum ferritin level: determinant of iron requirement in haemodialysis patients. JAMA 1977; 238: 601. 3 Bell AD, Kincaid WR, Morgan RG, et al Serum ferritin assay and bone-marrow iron stores in patients on maintenance hemodialysis. Kidney Int 1980; 12: 237. 4 Cotterill AM, Flather JN, Cattell WR, Barnett MD, Baker LRI. Serum ferritin concentration and oral iron treatment in patients on regular haemodialysis. Br Med J
1979; i: 790-91. 5 Jacobs A, Miller F, Worwood M 6
DANIEL B. MOWREY
Department of Psychology, Brigham Young University, Provo, Utah DENNIS E. CLAYSON
Department of Psychology, Mount
Union U.S.A.
et al. Ferritin in the serum of normal subjects and deficiency and iron overload. Br Med J 1972; iv: 206-08. Lipschitz DA, Cook JD, Finch CA A clinical evaluation of serum ferritin as an index of iron stores N Engl J Med 1974, 290: 1213-16. iron
7. Letsky EA, Miller F, Worwood M, et al. Serum ferritin in children with thalassaemia regularly transfused. J Clin Pathol 1974; 27: 652. 8 Beamish MR, Walter R, Miller F, et al. Transferrin iron, chelatable iron and ferritin in idiopathic haemochromatosis. Br J Haematol 1974; 27: 219-28. 9 Ali M, Fayemi AO, Rigolosi R, Frascino J, Marsden T, Malcolm D. Hemosiderosis in hemodialysis patients: an autopsy study of 50 cases. JAMA 1980; 244: 343-45.
College, Alliance, Ohio,
The effects of the powdered rhizome of Zingiber officinale on the symptoms of motion sickness were compared with those of dimenhydrinate and placebo in 36 undergraduate men and women who reported very high susceptibility to motion sickness. Motion sickness was induced by placing the blindfolded subject in a tilted rotating chair. Measurements of perceived degree of gastrointestinal distress were reported every 15 s for up to 6 minutes by means of psychophysical methods. Z. officinale was superior to dimenhydrinate in reducing motion sickness.
Summary
Introduction Lewis and Lewis,1 in their review of the therapeutic properties of plants, included the fluid extract of the rhizome of ginger (Zingiber officinale) among the natural products which mitigate symptoms of gastrointestinal distress, thus 2 continuing a tradition that dates back at least as far as 1597.2 Traditionally, investigations of ginger root have used fluid extracts.3 Our preliminary study, however, suggested that powdered whole root may have pronounced therapeutic effects. The purpose of our study was to determine whether the powdered whole root was effective in suppressing the gastrointestinal symptoms related to motion sickness. We used the psychophysical techniques of Stevens,4 who found that, when subjects were asked to indicate with
Fayemi AO, Braun EV, Malcolm D, Laraia S. Dissociation between hepatosplenic and marrow iron in liver cirrhosis. Arch Pathol Lab Med (in press). Isaacson C, Bothwell TH. Synovial iron deposits in black subjects with iron overload.
10. Ali M, 11.
Arch Pathol Lab Med 1981; 105: 487-89. 12.
We thank Gertrude Martin for the preparation of the manuscript and Mary Higgins Fougere for assistance in data collection. This study was supported by a grant from the Hemodialysis Foundation, Teaneck, N.J. Correspondence should be addressed to M. A., Director, Division of Pathology, Immunology, and Laboratories, Holy Name Hospital, 718 Teaneck Road, Teaneck, New Jersey 07666, U.S.A.
patients with
MOTION SICKNESS, GINGER, AND PSYCHOPHYSICS
Valberg LS, Simon JB, Manley PN, Corbett WE, Ludwig J. Distribution of storage iron as body stores expand in patients with hemochromatosis. J Lab Clin Med 1975; 86: 479-89
13. Heilmeyer L Symposium.
Human hyposideraemia in iron metabolism An International Berlin: Springer Verlag, 1964: 68: 202-05. 14. Ali M, Fayemi AO, Laraia S, Kasper V. Dissociation between stainable marrow and liver iron following iron-dextran therapy. NJ Med Soc (in press). 15. Lottenberg R, Kitchens CS, Roessler GS, Noyes WD. Iron studies in hemophilia. Arch Pathol Lab Med 1981; 105: 655-58. 16. Buchanan GR, Moore GC. Pulmonary hemosiderosis and immune thrombocytopenia. JAMA 1981; 246: 861-64. 17. Crichton RR. Ferritin: structure, synthesis and function. N Engl J Med 1971; 284: 1413. 18. Harrison PM, Hoare RJ, Hoy TG, et al. Ferritin and haemosiderin: structure and function. In: Jacobs A, Worwood M, eds Iron in biochemistry and medicine. London: Academic Press, 1974: 73-114. 19. Addison GM, Beamish MR, Hales CN, et al. An immunoradiometric assay for ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. J Clin Pathol 1972; 25: 326-29. 20. Eschbach JW, Cook JD, Scribner BH, Finch A. Iron balance in hemodialysis patients. Ann Intern Med 1977; 87: 710-13. 21. Reissman KR, Dietrich MR On the presence of ferritin in the peripheral blood of patients with hepatocellular disease. J Clin Invest 1956; 35: 588-95. 22. Jones PAE, Miller FM, Worwood M, et al. Ferritinaemia in leukaemia and Hodgkin’s disease Br J Cancer 1973; 27: 212-17. 23. Shoden A, Ing K, Sturgeon P. Iron storage IV: cellular distribution of excess liver iron. Am J Pathol 1962; 40: 671-79. 24. Shoden A, Ing K, Sturgeon P. Iron storage I. the influence of time on the redistribution of excess storage iron. Am J Pathol 1958; 34: 1139-47. 25. Wood JK, Milner PFA, Pathak UN. The metabolism of iron-dextran given as a totaldose infusion to iron deficient Jamaican subjects. Br J Haematol 1968; 14: 119-29. 26. Grimes AH, Hutt MSR. Metabolism of Fe5 9 dextran complex in human subjects Br Med J 1957; ii: 1074-77.
MAJID ALI A. OLUSEGUN FAYEMI JOSEPH FRASCINO
ROBERT RIGOLOSI EVALYNNE V. BRAUN ROBERTO SINGER
Departments of Pathology of College of Physicians and Surgeons of Columbia University, The Mount Sinai School of Medicine, New York, and Holy Name Hospital, and the Regional Hemodialysis Center, Holy Name Hospital, Teaneck, New Jersey, U.S.A. The relation between serum ferritin levels and stainable-iron deposits in the liver, bone marrow was investigated in 36 patients with and spleen, chronic renal failure who died after being on haemodialysis for 1-103 months. Elemental iron (mean, 5450 mg) had been given intravenously as iron dextran to patients in a long-term subgroup, who had been on dialysis for more than 3 months. The results of semiquantitative histochemical assessment of tissue iron in slides obtained at necropsy (scale 0 to 4+) were confirmed by chemical analyses of tissue iron. Serum ferritin levels correlated well with the degree of hepatosplenic siderosis but did not always correlate with bone-marrow iron stores in these patients. Serum ferritin concentrations were raised in 10 marrow-iron-depleted subjects (mean, 1336
Summary
ng/dl). The paradoxical association of hepatosplenic siderosis with iron depletion was observed in most of the patients on dialysis for less than 40 months. The histochemical data show that the bulk of intravenously injected iron dextran is taken up by the liver and spleen; that the hepatosplenic stores fail to be mobilised to the bone marrow; and that intravenous iron-dextran therapy, by-passing the intestinal mechanism for the regulation of iron absorption, carries a high risk of marrow
long-term hepatosplenic siderosis. Introduction
THE assay of serum ferritin concentrations to determine the requirements of haemodialysis patients has gained widespread acceptancel-4 on the basis that there is a iron
correlation between serum ferritin levels and bone-marrow iron stores. 1-3 Serum ferritin concentrations generally reflect total-body iron stores: low serum ferritin levels indicate iron deficiency,S,6 and the levels in disease states associated with tissue iron overload are high, sometimes over 5000 ng/dl. 6—8 The paradoxical coexistence of hepatosplenic siderosis and bone-marrow iron depletion has been observed in some patients with haemodialysis haemosiderosis9 and cirrhosis10, 11and in a single case each ofhaemochromatosis, 12 congenital atransferrinaemia,13 and anorexia nervosa.14 There have been reports of iron-depleted bone marrow in the presence of massive iron overload in other tissues in patients with haemophilials and pulmonary haemosiderosis.16 In patients with hepatosplenic siderosis, serum ferritin levels could theoretically be raised because of the elevated hepatosplenic iron stores even though the bone marrow is depleted of iron. This possibility has not yet been investigated with histochemical and biochemical studies. We now report data on serum ferritin and reticuloendothelial iron stores, marrow and hepatosplenic, in 36 the 13th annual meeting of the American Washington D.C., in 1980.
*Presented in part
Nephrology,
at
Society of
chronic renal failure who died after 1 to 103 months on maintenance haemodialysis. We have examined the relation between the duration of haemodialysis and the tissue iron stores in these and 19 other dialysis patients in order to clarify how intravenously administered iron dextran is used.
patients with
Patients and Methods We selected for the study 36 haemodialysis patients for whom data ferritin levels were available and whose necropsies were carried out consecutively at the regional haemodialysis centre at Holy Name Hospital, Teaneck, New Jersey. The results of an additional 19 consecutive necropsies on haemodialysis patients were included in an examination of the relation between tissue iron levels and the duration of dialysis. Clinical information including the age and sex of patients, duration of haemodialysis, quantity of intravenous iron dextran given, volume of packed cells and whole blood transfused, and other haematological data, was obtained from patients’ records. The raised serum ferritin levels had been established in most of the patients by two or more determinations separated by intervals of 1 month carried out when the patients were not receiving any blood transfusions. With few exceptions, the patients in the study had not received any iron dextran in the 6 months before death. 7 of the 36 patients had been on maintenance haemodialysis for less than 3 months and served as a control group. on serum
Serum Ferritin Serum ferritin
assayed with a double-antibody precipitation radioimmunoassay (Calbiochem-Behring kit). The mean serum ferritin levels measured with this technique in our laboratory for 20 normal volunteers and non-dialysis patients without haemotological disorders were for men, 186 ng/dl (range, 34-371 ng/dl) and for women, 117 ng/dl (range, 17-265 ng/dl). was
Morphological Studies We re-examined the routine histological sections which had been taken at necropsy, fixed with 4% buffered formaldehyde, and stained with haematoxylin and eosin. Two to four slides each of the liver, spleen, and bone marrow were available. Slides from all patients were stained with histochemical stains which use the prussian-blue reaction to detect iron. Known positive and negative control sections for iron preparations were stained with each batch of study slides. Sections of bone marrow from 3 non-dialysis patients who died without any known renal disease or disorder of iron metabolism were used for comparison when evaluating the bone-marrow sections of patients in the dialysis study group. Trichrome and reticulin stains of liver sections were studied to evaluate the reticular framework in 10 patients with advanced siderosis.
Tissue Iron Scores The iron stores in the marrow, liver, and spleen were evaluated according to a previously described9 semiquantitative system of scoring on a scale’of0 to 4 + (fig. 1). Diffuse blue stippling and fine blue granules were scored as 1 +; coarse heavy granules, sometimes coalescing to form large massive granules, were scored as 4+; and the blue granularity ranging between these two extremes was divided into moderate deposits scored as 2 + and heavier deposits scored as 3 +. Sections with no staining or rare minute foci of blue granularity were classified as negative. The marrow sections from the 3 non-dialysis patients were assigned a value of 2+ for comparison with marrow sections from dialysis patients. Iron stores found in areas of recent or old haemorrhage in these organs were excluded from this evaluation.
Quantitative Chemical Analysis of Tissue Iron All tissues for histology were fixed with neutral buffered 4% formaldehyde. The bone specimens were decalcified and fixed with a formic acid/formaldehyde solution for 24 h. To ensure that iron
653
of liver from 4 haemodialysis patients stained with prussian blue. (A). No stainable iron in liver of a patient on haemodialysis for 1 month only. (B) Iron deposits of score 1 +. (C). Iron stores representing 3+ score. (D). Massive iron deposits (4+) from a patient with advanced hepatic siderosis. Original magnification fields A and D, x40, fields B and C, x 100; all 4 are reduced by li’3.
Fig. I-Photomicrographs of sections
not dissolved from the tissue during fixation and decalcification, quantitative chemical analyses for iron were carried out on fresh and fixed tissues. For chemical analysis the air-dried tissue sample was weighed then digested with concentrated nitric acid, and the iron content in the sample was analysed by atomic absorption in a heated graphite furnace. The National Bureau of Standards’ standard no. 1577 (bovine liver, iron content 18 jug/g airdried weight) was used as the reference sample. was
Fig. 2-Correlation of semiquantitative scores haemodialysis patients. r=0’689: p
Results Clinical and Laboratory Data The 23 men and 13 women in the study died at age 46 to 76 years (mean age, 61’5years). The duration of maintenance haemodialysis ranged from 1 to 103 months; the mean for 29 patients in the long-term subgroup (>3 months) was 39 months. Negative-pressure dialysers of hollow-fibre and parallel-plate types were used for all patients from 1977 to 1979; some patients who had been on dialysis for more than 3 years had used coil-type dialysers during the preceding years. The patients in the long-term group received a mean of 9 -8 units of packed cells or whole blood during the period of treatment; blood was transfused either for control of a bleeding diathesis or when improvement in red-cell mass was clinically deemed urgent. All patients were given 100-105 mg/day iron orally as ferrous sulphate for the duration of the treatment. Patients on dialysis for more than 3 months received iron as iron dextran- (’Imferon’) intravenously (mean 5450 mg during the treatment period; range, 30014 850 mg). All 36 patients in the study were anaemic; the haemoglobin levels ranged from 5 -8 to 10 -4 g/dl (mean, 7’ 9 g/dl). Of 10 patients with marrow iron depletion in the presence of hepatosplenic siderosis, anaemia was microcytic in 8 patients and normocytic in 2; niild to moderate poikilocytosis was observed in all patients. ’
Hepatosplenic Siderosis Advanced hepatosplenic siderosis (3 + to 4 + iron deposits) was observed in 23 of the 36 patients (fig. 2). In the control group (<3 months on haemodialysis) there was no stainable iron in 3 patients and focal bluish stippling of the cytoplasm of the hepatocytes in 3 patients. In the liver iron deposits appeared first mainly in the Kupffer cells; in advanced siderosis there were coarse blue granules, at times coalescing to form large masses, within both the reticuloendothelial and
serum
for
ferritin
hepatosplenic
concentrations iron
deposits
with in 36
the hepatic parenchymal cells. In siderosis of the spleen, as in the liver, cells lining the sinusoids had the heaviest iron
deposits; malpighian corpuscles were generally spared even in advanced iron overload. Other notable histological features in the liver were mild fatty changes, central venous congestion, and an increase in the fibroconnective framework; the last was most prominent in advanced hepatic siderosis. Bone-marrow Iron Scores The distribution of marrow iron scores is shown in fig. 3. The marrow was iron-depleted (scored as negative) in 10 patients with 3+ to 4+ hepatosplenic siderosis and high serum ferritin concentrations (537-3994 ng/dl) (r=0-42;
p<0-01). Chemical Analysis
of Tissue Iron Semiquantitative histochemical estimation of tissue iron was validated by chemical analysis of iron in the liver, spleen, and bone marrow (see table). Fixed and decalcified samples of bone marrow showed larger quantities of iron than fresh samples (Sample 1: fresh bone 425
washed
However, bone-marrow iron does not appear to be any significant degree by this process.
out to
Serum Ferritin Levels ferritin concentrations correlated well with the semiquantitative iron scores for hepatosplenic iron deposits
The
serum
654
and bone-marrow iron deposits after intravenous iron-dextran therapy in 55 haemodialysis patients. Numbers in parentheses are the numbers of patients in each 10-month
Fig. 4-Hepatosplenic therapy period.
(fig. 2; r
diminished marrow-iron stores in the first 30 months of maintenance dialysis; in several patients the bone marrow was depleted. The 7 patients on dialysis less than 3 months had less iron in their bone-marrow sections than the control patients. During the first few months of dialysis the dichotomy between the hepatosplenic and the bone-marrow iron scores progressed until the paradox of hepatosplenic siderosis coexisting with marrow-iron depletion was reached in 21 patients. In many patients on dialysis for more than 30 months there were progressive increases in the stainable-iron content of the bone marrow, presumably after liver and spleen were saturated with iron. The 5 patients on dialysis for more than 80 months had massive iron deposits in the bone
did
marrow.
Fig. 3-Distribution of bone-marrow iron scores. = 0’ 689, p<0’ 001). By contrast, serum ferritin levels always correlate with the stainable iron scores in the marrow (fig. 3; r=0’42, p<0’01). Raised serum ferritin
not
found in - 10 marrow-iron-depleted 537-3994 ng/dl; mean 1336 ng/dl), in 6 of whom there was no stainable iron and in 4 there were rare minute deposits of iron. For 3 other patients with raised serum ferritin concentrations (1490-2510 ng/dl) marrow iron was scored as 2+ (the value for non-dialysis control marrow sections). In general, serum ferritin levels were higher in patients who had been on haemodialysis therapy for longer periods of time; mean values for 29 patients treated for more than 3 months and 7 patients treated for less than 3 months were 1820 ng/dl and 338 ng/dl, respectively. concentrations
were
patients (range,
Tissue Iron and Duration
of Dialysis All patients on dialysis for more than 20 months had moderate to heavy iron deposits in liver and spleen sections; there was massive siderosis of these organs in all patients on dialysis for more than 40 months. Though semiquantitative, the hepatosplenic iron scores were directly related to the duration of therapy (fig. 4). In contrast, the patients showed CORRELATION BETWEEN HISTOCHEMICAL ESTIMATION OF TISSUE IRON AND QUANTITATIVE CHEMICAL ANALYSIS IN 3 PATIENTS WITH TISSUE IRON OVERLOAD
Histochemical tissue iron
weight.
scale, 0-4+. Chemical analysis results in pg/g dry
Discussion Ferritin is the principal iron-containing protein in the tissues, and with its insoluble derivative, haemosiderin, it accounts for most of the iron identified in histological sections with the prussian-blue reaction. The ferritin molecule is thought to consist of a crystalline ferric-oxidephosphate core and a protein shell with a molecular weight of 450 000.’ Although many tissues can synthesise ferritin, the liver, spleen, and bone marrow are the primary sites of production. 17,11 Since an immunoradiometric procedure for its measurement in serum has been available 19 serum ferritin has been superior to the serum iron and serum iron-bindingcapacity tests as an indicator of marrow-iron reserves.l,3,. Serum ferritin concentrations below the normal range indicate iron deficiency;5,6,2o raised levels have been observed in disease states associated with increased body iron stores.6-8 Serum ferritin levels are raised in acute and chronic liver diseases and in some malignant disorders,21,22 as well as in disorders of iron metabolism. Our data agree with studies correlating serum ferritin concentrations and total-body iron stores in non-uraemic patients 17-20 but do not fully accord with some studies showing correlation between the bone-marrow iron and serum ferritin in haemodialysis patients. 1-3 In our study, when serum ferritin was low, the marrow was iron-depleted. but the converse did not always obtain. High serum ferritin levels in our patients with hepatosplenic siderosis were not correlated with the bone-marrow iron stores (fig. 3). In 10 patients, the serum-ferritin concentrations were high and yet there was little or no stainable iron in the bone marrow. In another 3 patients with high serum ferritin levels
655
(1490-2510 ng/dl) the marrow iron score was only 2 + (nondialysis control value). 1-3 In previous studies of serum ferritin and marrow iron, no attempt was made to investigate iron overload in the liver and spleen. In the absence ofhepatosplenic siderosis the marrowiron stores can be expected to represent total-body iron stores, and iron-depleted marrow can be regarded as evidence of iron deficiency. This is not true of dialysis haemosiderosis, where there may be a paradox of hepatosplenic iron overload and marrow iron depletion. In this setting, high serum ferritin levels are caused by hepatosplenic siderosis and may not be relied upon as indicators of marrow iron reserves. All our patients except those in the control group were given relatively large doses of iron dextran, whereas none of the patients in the study of Bell et al. received parenteral iron. Another possible cause of the different results in our study and that of Bell et awl. is that the tissue obtained by needle biopsy of bone is usually scanty and crushed and does not allow as accurate an estimate of marrow iron as do large sections of properly prepared bone obtained at necropsy. The solution used to decalcify the bone-marrow sections may cause underestimation of marrow iron, but identical histochemical procedures were used for the dialysis patients’ and controls’ sections and the results of chemical analyses of tissue iron agreed with the histochemical scores (table). The fate ofhepatosplenic iron stores after intravenous irondextran therapy has not been fully investigated. In rabbits the bulk of intravenously injected iron dextran is taken up by liver and spleen;23,24 when the animals were killed 6 months after iron-dextran overload, even larger quantities of iron were found in the liver. 24 After intravenous injection of radiolabelled iron dextran in man, more radioactivity was found in the liver than in any other organ at the end of the study period (4 to 6 weeks).25,26 Our study shows that there is a failure of mobilisation of hepatosplenic iron to the bone marrow in this setting and that hepatosplenic iron overload may persist even in the face of bone-marrow iron depletion. Our data raise serious questions about the appropriateness of intravenous iron-dextran therapy for treatment of iron deficiency in dialysis patients. They show that marrow iron cannot be replenished by intravenous iron dextran without incurring serious hepatic iron overload and the risk of liver
injury.
REFERENCES
1. Hussein S, Prieto J, O’Shea M, Hoffbrand AV, Baillod RA, Moorhead JF. Serum ferritin assay and iron status in chronic renal failure and haemodialysis. Br Med J 1975,i: 546-48. 2 Mirahmadi SM, Paul WL, Winer RL, et al Serum ferritin level: determinant of iron requirement in haemodialysis patients. JAMA 1977; 238: 601. 3 Bell AD, Kincaid WR, Morgan RG, et al Serum ferritin assay and bone-marrow iron stores in patients on maintenance hemodialysis. Kidney Int 1980; 12: 237. 4 Cotterill AM, Flather JN, Cattell WR, Barnett MD, Baker LRI. Serum ferritin concentration and oral iron treatment in patients on regular haemodialysis. Br Med J
1979; i: 790-91. 5 Jacobs A, Miller F, Worwood M 6
DANIEL B. MOWREY
Department of Psychology, Brigham Young University, Provo, Utah DENNIS E. CLAYSON
Department of Psychology, Mount
Union U.S.A.
et al. Ferritin in the serum of normal subjects and deficiency and iron overload. Br Med J 1972; iv: 206-08. Lipschitz DA, Cook JD, Finch CA A clinical evaluation of serum ferritin as an index of iron stores N Engl J Med 1974, 290: 1213-16. iron
7. Letsky EA, Miller F, Worwood M, et al. Serum ferritin in children with thalassaemia regularly transfused. J Clin Pathol 1974; 27: 652. 8 Beamish MR, Walter R, Miller F, et al. Transferrin iron, chelatable iron and ferritin in idiopathic haemochromatosis. Br J Haematol 1974; 27: 219-28. 9 Ali M, Fayemi AO, Rigolosi R, Frascino J, Marsden T, Malcolm D. Hemosiderosis in hemodialysis patients: an autopsy study of 50 cases. JAMA 1980; 244: 343-45.
College, Alliance, Ohio,
The effects of the powdered rhizome of Zingiber officinale on the symptoms of motion sickness were compared with those of dimenhydrinate and placebo in 36 undergraduate men and women who reported very high susceptibility to motion sickness. Motion sickness was induced by placing the blindfolded subject in a tilted rotating chair. Measurements of perceived degree of gastrointestinal distress were reported every 15 s for up to 6 minutes by means of psychophysical methods. Z. officinale was superior to dimenhydrinate in reducing motion sickness.
Summary
Introduction Lewis and Lewis,1 in their review of the therapeutic properties of plants, included the fluid extract of the rhizome of ginger (Zingiber officinale) among the natural products which mitigate symptoms of gastrointestinal distress, thus 2 continuing a tradition that dates back at least as far as 1597.2 Traditionally, investigations of ginger root have used fluid extracts.3 Our preliminary study, however, suggested that powdered whole root may have pronounced therapeutic effects. The purpose of our study was to determine whether the powdered whole root was effective in suppressing the gastrointestinal symptoms related to motion sickness. We used the psychophysical techniques of Stevens,4 who found that, when subjects were asked to indicate with
Fayemi AO, Braun EV, Malcolm D, Laraia S. Dissociation between hepatosplenic and marrow iron in liver cirrhosis. Arch Pathol Lab Med (in press). Isaacson C, Bothwell TH. Synovial iron deposits in black subjects with iron overload.
10. Ali M, 11.
Arch Pathol Lab Med 1981; 105: 487-89. 12.
We thank Gertrude Martin for the preparation of the manuscript and Mary Higgins Fougere for assistance in data collection. This study was supported by a grant from the Hemodialysis Foundation, Teaneck, N.J. Correspondence should be addressed to M. A., Director, Division of Pathology, Immunology, and Laboratories, Holy Name Hospital, 718 Teaneck Road, Teaneck, New Jersey 07666, U.S.A.
patients with
MOTION SICKNESS, GINGER, AND PSYCHOPHYSICS
Valberg LS, Simon JB, Manley PN, Corbett WE, Ludwig J. Distribution of storage iron as body stores expand in patients with hemochromatosis. J Lab Clin Med 1975; 86: 479-89
13. Heilmeyer L Symposium.
Human hyposideraemia in iron metabolism An International Berlin: Springer Verlag, 1964: 68: 202-05. 14. Ali M, Fayemi AO, Laraia S, Kasper V. Dissociation between stainable marrow and liver iron following iron-dextran therapy. NJ Med Soc (in press). 15. Lottenberg R, Kitchens CS, Roessler GS, Noyes WD. Iron studies in hemophilia. Arch Pathol Lab Med 1981; 105: 655-58. 16. Buchanan GR, Moore GC. Pulmonary hemosiderosis and immune thrombocytopenia. JAMA 1981; 246: 861-64. 17. Crichton RR. Ferritin: structure, synthesis and function. N Engl J Med 1971; 284: 1413. 18. Harrison PM, Hoare RJ, Hoy TG, et al. Ferritin and haemosiderin: structure and function. In: Jacobs A, Worwood M, eds Iron in biochemistry and medicine. London: Academic Press, 1974: 73-114. 19. Addison GM, Beamish MR, Hales CN, et al. An immunoradiometric assay for ferritin in the serum of normal subjects and patients with iron deficiency and iron overload. J Clin Pathol 1972; 25: 326-29. 20. Eschbach JW, Cook JD, Scribner BH, Finch A. Iron balance in hemodialysis patients. Ann Intern Med 1977; 87: 710-13. 21. Reissman KR, Dietrich MR On the presence of ferritin in the peripheral blood of patients with hepatocellular disease. J Clin Invest 1956; 35: 588-95. 22. Jones PAE, Miller FM, Worwood M, et al. Ferritinaemia in leukaemia and Hodgkin’s disease Br J Cancer 1973; 27: 212-17. 23. Shoden A, Ing K, Sturgeon P. Iron storage IV: cellular distribution of excess liver iron. Am J Pathol 1962; 40: 671-79. 24. Shoden A, Ing K, Sturgeon P. Iron storage I. the influence of time on the redistribution of excess storage iron. Am J Pathol 1958; 34: 1139-47. 25. Wood JK, Milner PFA, Pathak UN. The metabolism of iron-dextran given as a totaldose infusion to iron deficient Jamaican subjects. Br J Haematol 1968; 14: 119-29. 26. Grimes AH, Hutt MSR. Metabolism of Fe5 9 dextran complex in human subjects Br Med J 1957; ii: 1074-77.