-
METABOLISM OF IRON I N HEMOCHROMATOSIS*
MILTON R. BEYERS, M.D., AND STANLEY E. GITLOW, M.D.t
From the Department of Internal Medicine, Veterans Administration
Kingsbridge Road, Bronx, New York
Hospital,
Hemochromatosis is characterized by extensive deposition of iron throughout
the body, particularly in the liver, pancreas and skin, and by the clinical triad
of hepatic cirrhosis, diabetes and pigmentation of the skin. Normally, the total
amount of iron in the body, including the blood, varies from.3 to 5 grams, whereas
in hemochromatosis the liver alone may contain more than 20 grams. The total
iron in the tissues, exclusive of blood, has been found to be over 50 grams in
some instances. 2, 6 Sheldon has reported that the iron content of the liver in
"hemochromatosis varies from 2.1 to 3.6 Gm. per 100 Gm. of liver tissue as compared with the^normal average of 0.05 Gm. per 100 Gm.16
Most authors agree and the majority of clinical and experimental studies
point to the fact that the deposition of iron antecedes the characteristic widespread fibrosis seen in hemochromatosis.1' 5> 6 ' 10"12, l6, 16T I8, l9 This does not,
however, signify that the pigment per se is the sole cause of the ensuing fibrosis
of the organ. It is quite conceivable that excessive deposition of hemosiderin
acts as an initial insult allowing other toxic factors, such as vitamin deficiencies,
viral hepatitis, and generalized infections to produce the fibrosis. This initial
insult may be caused by an inhibition of an essential enzyme system.13
Assuming that deposition of iron is the factor that induces fibrosis which in
turn results in the cirrhosis and diabetes in hemochromatosis, removal of the
excess of stored iron shauld be of therapeutic value. Inasmuch as the diet is the "
ultimate source of tHe excess of iron, and because it is generally accepted at the
present tHj^that the bodily excretion of iron is minimal, it becomes apparent
that,there are two major methods by which the iron content of the body may
be'.'decreased: one is control of the source of iron in the diet, and the other is
withdrawal of blood, the tissue that is rich in iron. Both of these methods have
been suggested in the treatment of hemochromatosis. The primary assumption for advocating phlebotomy is that the patient with hemochromatosis is
able under an appropriate stimulus to utilize his excessive iron stores for the
production of hemoglobin. To the best of our knowledge, this assumption has
never been conclusively proven. Finch4 and Rundles14 failed to observe the
development of anemia following repeated phlebotomies in patients with hemochromatosis who were on normal diets. This would seem to indicate that such
* Reviewed in the Veterans Administration and published with the approval of the Chief
Medical Director. The statements and conclusions published by the authors are a result of
their own studies and do not necessarily reflect the opinion or policy of the Veterans Administration.
Received for publication, October 30, 1950.
f Present address, Department of Pharmacology-Physiology, University of the State of
New York, College of Medicine, Brooklyn, New York.
349
350
BEYERS AND GITLOW
patients are able to utilize their excess of stored iron for the production of
hemoglobin. However, without control of the dietary iron, such a conclusion
is only presumptive.
The utilization of stored iron in hemochromatosis may be quantitatively
evaluated by repeatedly removing large quantities of hemoglobin in such a
patient, maintaining an iron-free diet and observing the degree of regeneration
of hemoglobin. If, on such a regimen, regeneration of hemoglobin occurs, it is
a priori evidence that the patient has utilized his iron stores. We were afforded
the opportunity of conducting such a study in a young patient with proven
hemochromatosis, whose physical condition permitted the rigors of this regimen.
METHOD
A 27-year-old white man, weighing GO Kg., with hemochromatosis proven by
biopsy of liver and skin, was maintained throughout the study, except for an
initial seven-day control period, on a diet calculated to provide: (1) a total of
2.5 mg. of iron or less daily; (2) 1400 to 1600 calories a day; and (3) 50 to 60
Gm. of protein daily. This diet was supplemented by the daily oral administration of 10 mg. of thiamine, 25 mg. of nicotinic acid, 5 mg. of riboflavin, and 100
mg. of ascorbic acid, and weekly intramuscular injections of 15 /ig. of crystalline vitamin B 12 (Cobione, Merck and Co., Rahway, N. J.). It was felt that this
diet provided all the necessary hemoglobin precursors except adequate amounts
of iron. The total iron content of one day's sample diet was found to be 2.7 mg.
The patient drank only iron-free distilled water (less than one part iron in a
million) and all cooking was done in aluminum utensils using iron-free distilled
water.
Control values for hemoglobin, erythrocyte count, hematocrit reading and
plasma volume were determined during the initial seven-day period. In the
next 36 days of the study, 13 phlebotomies were performed, the amount of blood
withdrawn varying from 175 to 490 ml. Following an interval of 30 days, an
additional 1500 ml. of blood Avas removed from the patient and he was again
observed for 18 days.
Serial determinations of hemoglobin, erythrocyte count, hematocrit reading,
reticulocyte count and plasma A'olume were performed throughout the study.
The hemoglobin determinations were performed four to five times a week by
the same technician, using a Klett-Summerson photoelectric colorimeter. Two
simultaneous determinations on two different fingers for 25 successive days established the average error for this method as ±0.22 grams per 100 ml.
The hematocrit reading was obtained simultaneously with the hemoglobin
determinations using the technic outlined by Wintrobe.17 The accepted error
for this method is ± 0 . 5 per cent.
The Evans blue dye (T1824) method of determining plasma volume was used
at intervals of seven to twelve days. Although the dye method for plasma volume
determination is not reliable for the calculation of an absolute value, it is, nevertheless, quite accurate in performing serial determinations for comparison in the
same person. In this laboratory, consecutive determinations in a large series of
IRON METABOLISM IN HEMOCHROMATOSIS
351
patients have shown a maximum error of ±200 ml. The variation between two
control values for this patient was 180 ml.
The total circulating hemoglobin was calculated by the formula:
Plasma volume X hemoglobin (Gm. per 100 ml.)
100 — hematocrit reading
= total circulating hemoglobin (Gm.)
Using this formula and noting the previously listed errors, the maximum total
circulating hemoglobin error was calculated as ± 7 0 Gm.
One clay's sample diet, which we calculated to contain 2.5 mg. of iron, was
analyzed by the method of Farrar 3 and found to contain 2.7 mg. of iron. We
consider the difference of 0.2 mg. to be Avithin our range of error since the diet
Avas undoubtedly contaminated to a slight degree during the long process of
desiccation.
RESULTS
All pertinent experimental data are presented in graphic form in Figures 1
and 2. For purpose of discussion, the total study is diAnded into fiVe periods:
Control period (clays 1 through 7), first phlebotomy period (days 8 through 43),
first observation period (days 44 through 73), second phlebotomy period (days
74 and 75), and second observation period (clays 76 through 93).
/ . Control period (days 1 through 7). Successive hemoglobin determinations were
14.8, 14.6, and 14.7 grams per 100 ml.; repeated hematocrit readings gaAre a
value of 42 per cent; and the reticulocyte count was 1.2 per cent.
Plasma volume determinations at a two-day interval were 3240 ml. and 3060
ml. Avith a resultant total circulating hemoglobin of 827 Gm. and 786 Gm.,
averaging 807 Gm. During this period, the patient Avas on a normal diet and
his body weight Avas 132 pounds (60 Kg.).
Liver function studies including cephalin flocculation, total protein, albuminglobulin ratio, alkaline phosphatase, serum bilirubin, thymol turbidity, and
bromsulfalein excretion were within normal limits. Urinary urobilinogen excretion
was consistently elevated. The glucose tolerance Avas slightly diminished but
the initial fasting blood sugar level was 109 mg. per 100 ml.
II. First phlebotomy period (days 8 through 48). At the beginning of this period,
the patient was placed on the special iron-deficient diet and this was continued
for the remainder of the study. A total of 4190 ml. of blood was withdrawn.
This contained a calculated 514 Gm. of hemoglobin, equivalent to 1721 mg. of
iron (hemoglobin in Gm. X 3.35 = mg. of iron).2
The hemoglobin and hematocrit A'alues diminished progressiA'ely. The former
reached its lowest A'alue of 9.1 Gm. per 100 ml. on the thirty-eighth day and
the latter fell to 27 per cent at the end of the period. It may be noted that the
hemoglobin A'alue increased from 9.1 to 11.4 Gm. per 100 ml. between the thirtyeighth and forty-third days, during which time the phlebotomies were less frequent. An increased reticulocyte count, 2.3 per cent, was first noted on the eighteenth clay and it steadily rose to 6.2 per cent on the thirty-ninth day.
352
BEYERS AND GITLOW
The total circulating hemoglobin fell progressively and reached 450 Gm. by
the end of this period. Although his general condition remained excellent, the
patient's weight decreased to 122 pounds (55.4 Kg.).
III. First observation period (days 44 through 73). The hemoglobin concentration continued to fall, reaching 10.8 Gm. per 100 ml. on the forty-sixth day;
thereafter it rose progressively to 14.0 Gm. per 100 ml. by the seventy-third
day. The hematocrit value increased gradually to 40 per cent by the end of
D A Y S
FIG. 1. Hematocrit, hemoglobin and reticulocyte values during the study
this period. Reticulocyte counts remained elevated until the forty-eighth day
and then rapidly fell, reaching 1 per cent on the fifty-fourth day.
The total circulating hemoglobin reached the lowest value, 415 Gm. on the
forty-eighth day, representing a loss of 392 Gm. when compared with the average control value. Thereafter, the total circulating hemoglobin rose steadily to
reach 845 Gm. by the end of this observation period, an increase of 430 Gm.
during a 26-day interval.
The discrepancy between the calculated loss of 514 Gm. of hemoglobin in
the blood removed during the first phlebotomy period, and the observed loss of
392 Gm. of hemoglobin, as determined by the total circulating hemoglobin
values, was probably the result of continued hemoglobin production during
IRON METABOLISM IN HEMOCHROMATOSIS
353
the phlebotomy period. In the second phlebotomy period, limited to two days,
this discrepancy was not noted.
Repeated liver function studies gave results which were within normal limits;
the fasting blood sugar on the sixtieth day was 123 mg. per 100 ml., but repeated
urine examinations for glycosuria were negative. The patient's weight remained
122 pounds (55.4 Kg.).
4CO-l-T-T-1-TTT-r
UJ
WOO-
D A Y S
FIG. 2. Plasma volume and total circulating hemoglobin values during the study
IV. Second phlebotomy period (days 7J, and 75). A total of 1500 ml. of blood
was withdrawn; this contained a calculated 197 Gm. of hemoglobin, equivalent
to 661 mg. of iron.
V. Second observation period (days 76 through 93). Hemoglobin values fell to
a low of 10.0 Gm. per 100 ml. on the eighty-first day. At this time the hematocrit
reading was also at its lowest level, 29 per cent. By the end of this period, the
hemoglobin had risen to 13.8 Gm. per 100 ml., and the hematocrit reading to
39.5 per cent. The reticulocytes gradually increased to 8.3 per cent on the eightyfourth day following which, they diminished to 1.0 per cent at the end of the
period.
354
BEYERS AND GITLOW
The total circulating hemoglobin diminished to 680 Gm. by the eighty-second
day, representing a loss of 165 Gm. of hemoglobin when compared with the
value just prior to the second phlebotomy period. It will be noted that the figure
of 165 Gm. of hemoglobin, as determined from the total circulating hemoglobin
values, compared favorably with the calculated loss of 197 Gm. of hemoglobin
in the 1500 ml. of blood removed during the second phlebotomy period. The
total circulating hemoglobin rose to reach 939 Gm. by the ninety-third day, an
increase of 259 Gm. in a 12-day interval.
Immediately after this period, liver function studies were again found to be
normal. A second glucose tolerance curve did not differ significantly from the
control study.
DISCUSSION
It has already been noted that in the first phlebotomy period, the calculated
loss of 514 Gm. of hemoglobin in the blood withdrawn failed to correlate with
the observed loss of 392 Gm. of hemoglobin as determined by comparison of
the total circulating hemoglobin values. This discrepancy was obviously caused
by the patient's continued synthesis of hemoglobin during the 36-day interval
of blood loss. The total circulating hemoglobin value at the end of this phelbotomy period represented the amount of hemoglobin remaining after the phelbotomies plus the amount regenerated during this interval. In the second phelbotomy
period, of only two days' duration, the calculated loss of 197 Gm. of hemoglobin
in the withdrawn blood more closely approximated the actual 165 Gm. difference
between the total circulating hemoglobin vames before and after the phlebotomies. It is, therefore, evident that comparisons of the total circulating hemoglobin values cannot be used in determining the absolute amount of hemoglobin
lost or regenerated, unless the blood loss occurred over a short period of
time and frequent determinations of the total circulating hemoglobin have been
made.
The only accurate method for determining the total amount of synthesized
hemoglobin is by adding the total calculated hemoglobin in the phlebotomized
blood to the difference between the initial and final total circulating hemoglobin levels [total synthesized hemoglobin = total hemoglobin in withdraAvn
blood + (final total circulating hemoglobin-initial total circulating hemoglobin)].
During the entire study, 711 grams of hemoglobin, equivalent to 2382 milligrams of iron, were removed by phlebotomy. The initial total circulating hemoglobin was 807 grams and the final value 939 grams, an increase of 132 grams.
The total synthesized hemoglobin was, therefore, 843 grams (132 + 711).
The calculated maximum error for total circulating hemoglobin was ± 7 0 Gm.
Assuming this error in both the initial and final total circulating hemoglobin
determinations, the maximum error in computing the total synthesized hemoglobin was ± 1 4 0 Gm. If this error were actually present the minimum total
synthesized hemoglobin would have been 703 grams (843 — 140).
This value of 703 grams of total synthesized hemoglobin can be divided into
three components: (1) that derived from dietary iron; (2) that derived from the
IRON METABOLISM IN HEMOCHROMATOSIS
355
iron stores normally available in man; and (3) that derived from iron stores in
excess of those found in normal persons.
During the 86 days of the special diet, the patient received a total of 215 mg.
iron from his food (86 X 2.5 mg. Fe per day). Part of this dietary iron was in
an unavailable form; part of the available iron was probably not absorbed; and
a portion of the diet offered was not eaten. However, assuming 100 per cent
availability, absorption and ingestion, the 215 mg. of dietary iron would have
enabled the patient to synthesize 64 Gm. of hemoglobin.
The normal total body iron is about 45 mg. per kilogram body weight, of
which 20 per cent is stored but available for hemoglobin synthesis.2 Therefore,
a normal person of the patient's weight (60 Kg.) would have 2700 mg. of total
iron, of which 540 mg. Avould be available for the production of 161 Gm. of
hemoglobin.
The third component of the total synthesized hemoglobin, that derived from
the iron stores in excess of those found in normal humans, is equal to the difference between the total synthesized hemoglobin and the sum of the first two
components listed above. These first two components (dietary iron, and normally
available storage iron) could account for 225 grams of the total 703 grams of
synthesized hemoglobin. The remaining 478 grams of hemoglobin must have
been produced from the excessive iron stores present in this patient. This represents the utilization of 1600 mg. of stored iron in excess of that amount ordinarily found in a person of the patient's weight. We feel that this-isdefinite proof
of the physiologic availability of the excess iron stores in this patient with hemochromatosis.
It is impossible to determine, on the basis of this study alone, the molecular
structure of stored iron which was mobilized for the production of hemoglobin.
The storage iron in the patient with hemochromatosis exists as hemosiderin and
as a high concentration of structurally normal ferritin, which contains iron of
the three unpaired electron type (three unpaired electrons in the outer electron
shell per atom of iron). 6, Si 9 Granick has stated that hemosiderin does not yield
utilizable iron as readily as ferritin. He implied, however, that hemosiderin iron
could be given up to the body for utilization slowly over a long period of time.7
Accepting this hypothesis, it may be assumed that the iron mobilized during
this relatively short study was derived for the most part from ferritin. Apparently
the stored iron in patients with hemochromatosis is normal in physiologic function as well as structure.
It is the consensus that excessive iron deposition is the primary factor in
causing the fibrosis found in hemochromatosis. Having now established in one
such patient that this excess iron is physiologically available for hemoglobin
regeneration, we suggest that an iron-deficient diet and repeated phlebotomies
offer a rational means of therapy.
SUMMARY
Blood was repeatedly withdrawn from a patient with hemochromatosis while
he was on an iron-deficient diet.
356
BEYERS AND GIT-LOW
By serial studies of the total circulating hemoglobin, it was observed that
this patient synthesized hemoglobin far in excess of that which would be expected
from normal stores of iron.
The physiologic availability of the excessive stores of iron for hemoglobin
production was demonstrated in a patient with hemochromatosis.
Acknowledgment.
We gratefully acknowledge t h e guidance of D r . Leo M . Meyer, t h e
assistance of Miss Constance H . Shine in preparing the diet, t h e technical assistance of M r .
Randolph S. Douglas, M r . Manuel J . Villazon, D r . Bernard Klein and M r . Milton Weissman, and t h e cooperation of Mrs. Clara J. H a r t , librarian.
Author'8 note. Since the completion of this study, Davis, W. D . , and Arrovvsmith, W. R .
(J. L a b . and Clin. Med., 3 6 : 814-815, 1950) have reported t h e therapeutic use of repeated
phlebotomies in patients with hemochromatosis a n d demonstrated histologic evidence of
depletion of the iron deposits in the liver.
REFERENCES
1. BOYD, W.: A Text-Book of Pathology. E d . 4. Philadelphia: Lea and Febiger, 1943, p ;
45.
2. CANTAROW, A., AND T R U M P E R , M . : Clinical Biochemistry. E d . 4. Philadelphia a n d London: W. B . Saunders Company, 1949, p p . 205-209.
3. FARRAR, C . E . , J R . : Determination of iron in biological materials. J . Biol. Chem., 110:
685-694, 1935.
4. F I N C H , C. A . : Iron metabolism in hemochromatosis. J . Clin. Investigation, 28: 780781, 1949.
5. G R A E F , I . , N E W M A N , W., G O R D O N , B . , AND O L I V E T T I ,
6.
7.
8.
9.
10.
11.
R . : O b s e r v a t i o n s on exogenous
hemochromatosis apparently d u e t o multiple transfusions. Read before t h e N . Y .
P a t h . S o c , M a y 25, 1950. P r o c N . Y. P a t h . S o c , in press.
GRANICK, S.: Iron metabolism and hemochromatosis. N . Y. Acad. Med., 25: 403-428,
1949.
GRANICK, S.: Discussion a t the N . Y . P a t h . S o c , M a y 25, 1950. Proc. N . Y. P a t h . S o c ,
in press.
GRANICK, S., AND H A H N , P . F . : F e r r i t i n ; speed of uptake of iron b y t h e liver and i t s
conversion t o ferritin iron. J . Biol. Chem., 155: 661-669, 1944.
GRANICK, S., AND MICHABLIS, L . : T h e presence of ferritin in t h e duodenal mucosa and
liver in hemochromatosis. Proc. Soc. Exper. Biol, and Med., 66: 296-298, 1947.
HOWARD, C. P . , AND M I L L S , E . S.: Oxford Medicine Vol. V I I I . New York: Oxford University Press, 1949, p p . 215-222.
K I N N E Y , T . D . , H E G S T E D , D . M., AND F I N C H , C. A.: T h e influence of diet on iron absorption. I . T h e pathology of iron excess. J . Exper. Med., 90: 137-145, 1949.
12. M U I R H E A D , E . E . , C R A S S , G., J O N E S , F . , AND H I L L , J . M . : I r o n overload (hemosiderosis)
aggravated b y blood transfusions. Arch. I n t . Med., 83: 477-501, 1949.
13. RACHER, E . , AND KRIMSKY, I . : Inhibition of coupled phosphorylation in brain homogenates by ferrous sulfate. J . Biol. Chem., 173: 519-533, 1948.'
14. R U N D L E S , W.: Personal communication to the authors.
15. SCHWARTZ, S. O., AND BLUMENTHAL, S. A.: Exogenous hemochromatosis resulting from
blood transfusions. Blood, 3: 617-640,1948.
16. SHELDON, J . H . : Haemochromatosis. New York and London: Oxford University Press,
1935, 382 p p .
17. W I N T R O B E , M. M . : Clinical Hematology. Ed. 2. Philadelphia: Lea a n d Febiger, 1946,
pp. 242-244.
IS. WYATT, J . B . , AND GOLDENBURC, H . : Hemosiderosis in refractory anemia. Arch. I n t .
Med., 83: 67-76, 1949.
19. ZELTMACHER, K., AND BEVANS, M . : Aplastic anemia and its association with hemochromatosis. Arch. I n t . Med., 75: 395-103, 1945.