The Use of Two Radioactive Isotopes of Iron in Tracer Studies of Erythrocytes

THE USE OF TWO RADIOACTIVE ISOTOPES OF IRON IN TRACER STUDIES OF ERYTHROCYTES

Wendell C. Peacock, … , Soma Weiss, John G. Gibson 2nd

J Clin Invest. 1946;25(4):605-615. https://doi.org/10.1172/JCI101743.

Research Article

Find the latest version: https://jci.me/101743/pdf THE USE OF TWO RADIOACTIVE ISOTOPES OF IRON IN TRACER STUDIES OF ERYTHROCYTES 1 By WENDELL C. PEACOCK, ROBLEY D. EVANS, JOHN W. IRVINE, JR., WILFRED M. GOOD, ARTHUR F. KIP, SOMA WEISS,2 AND JOHN G. GIBSON, 2ND (From the Radioactivity Center, Massachusetts Institute of Technology, Cambridge; the Medical Clinic of the Peter Bent Brigham Hospital, and Department of Medicine, Harvard Medical School, Boston) (Received for publication January 18, 1946)

INTRODUCTION Ross (6) used these mixed isotopes, but their There are 2 radioactive isotopes of iron suit- radioactivity detection apparatus was sensitive able for use as biological tracers. One of these, only to the radiations from Fe59. The nuclear Fe59, has a half life of 47 days and emits gamma reactions are Fe58 (d, p), Fe59 (Fe58, a stable rays and low energy beta rays. The other, Fe55, isotope, bombarded with deuterons, d, forms, has a half life of about 5 years and emits low through the liberation of a proton, p, Fe59) and energy x-rays only. Fe54 (d, p) Fe55. When iron is bombarded, This paper describes how these 2 isotopes are these 2 isotopes are produced simultaneously, and prepared for red blood cell tracer studies, how cannot be isolated thereafter in a state of radio- biological samples containing the radioactive iron active purity. are prepared for measurements, how the measure- By deuteron bombardment of manganese, the ments are performed, and how the results are reaction Mn55 (d, 2n) Fe55 occurs and only the reported to collaborating medical research teams long-lived radioactive iron isotope is obtained. engaged in various physiological and clinical The manganese targets are prepared from an al- studies. loy of manganese (90 per cent Mn, 8 per cent Cu, Fe59 has been in use for a number of years for 2 per cent Si) which has suitable mechanical and metabolism and red cell studies (1 to 5) and a thermal properties. This alloy is cast into %-inch method of preparing samples for measurement slabs and machined down to %2-inch thickness. with a thin window beta ray counter has been Targets 'A x 1 inch are cut from this material described by Ross (6). The longer lived iso- and silver soldered to the oscillating probe target tope, Fe55, on the other hand, appears not to described by Livingston (7). Evaporation of the have been used previously. target face limits the deuteron beam current to Each isotope has characteristics that make it 100 /La at 14 MEV. Under these conditions our particularly useful in certain problems; the com- cyclotron yields of Fe55 have been approximately bined use of both tracers offers solutions of % of a microcurie per microampere hour. problems which cannot be obtained by the employ- The shorter-lived isotope, Fe59, can be produced ment of either isotope alone. in pure form by the bombardment of cobalt with neutrons, through the nuclear reaction Co59 (n, p) Fe59. PREPARATION OF THE ISOTOPES Deuteron bombardment of cobalt also yields Fe59 as a by-product of the fast neutrons always a. Nuclear bombardment. Bombardment of present with deuteron reactions. Because thin iron with deuterons yields both isotopes mixed. cobalt targets are necessary for thermal conduc- The earlier work of Hahn et al. (1 to 5) and tion, and the total amount of cobalt is small (1 to 2 grams), this is not an efficient method of pro- 1 The work described in this paper was done under a ducing high Fe59 activities. If a comparatively contract, recommended by the Committee on Medical Re- large piece of cobalt (about 40 grams) is soft search, between the Office of Scientific Research and De- the of a velopment and the Massachusetts Institute of Technology, soldered to back beryllium target the in collaboration with the Peter Bent Brigham Hospital. larger amount of cobalt, higher permissible beam 2Deceased, January 31, 1942. currents on Be, and much higher neutron yield 605 606 PEACOCK, EVANS, IRVINE, GOOD, KIP, WEISS, AND GIBSON from the Be9 (d, n) BI0 reaction, give much better The final step in the preparation of the iron is yields of Fe59. Using beryllium, our yields have to synthesize it into a compound suitable for in- been approximately 0.023 microcuries per micro- travenous use in animals and humans. Ferric ampere hour. ammonium citrate, in doses described below, has -b. Radiochemical preparation. Regardless of been found safe for intravenous injection. After the mode of formation of the radioactive iron, the dissolving the Fe2O3 in concentrated HCI and purification and preparation of material for use is diluting, Fe(OH)3 is again precipitated with essentially the same in all cases. The target mate- NH4OH and is well washed with distilled water. rial Fe, Mn, or Co is dissolved in concentrated The precipitate is dissolved in a solution contain- hydrochloric acid, diluted and filtered. If man- ing 3 moles of citric acid for each mole of iron ganese and cobalt of high purity are used it is present. Warming the mixture hastens the solu- necessary to add 1 to 10 mgm. of inactive Fe+++ tion of the precipitate. When the precipitate is as a carrier. Usually there is enough iron in the completely dissolved the solution is brought to a target so that such additions are unnecessary. pH of 7.0 with NH4OH and filtered into a clean After oxidation of the iron from Fe++ to Fe+++ bottle. with H202 the iron is precipitated as Fe (OH)3 This solution can be diluted with neutral dis- with pyridine (8). tilled water to obtain any desired concentration The precipitate of Fe (OH)3 is dissolved in 6 M of iron. It withstands sterilization by autoclaving HC1, and 10 mgm. each of Mn, Co and Zn added if sealed in glass ampouls, but heating in open con- as carriers to aid in eliminating the radioactive iso- tainers may result in the formation of a precipi- topes of these elements from the iron. Ferric tate. Ferric ammonium citrate is photosensitive hydroxide is again precipitated with pyridine, and should be kept in the dark to avoid precipita- leaving the Mn, Co and Zn in solution. This tion. The compound does not inhibit the growth precipitation procedure is repeated 2 to 4 times of common molds. until the activity in the filtrate is low. After elimination of these 3 elements, the hy- PREPARATION OF RADIOACTIVE RED CELLS drogen sulphide group is eliminated by adding 50 When radioactive ferric ammonium citrate is' mgm. of Cu++ as carrier for this group, and pre- injected intravenously into animals or humans it cipitating the sulphide with H2S from a 0.1 M is rapidly removed from the blood stream. A HCl solution. The preciptate is filtered off and large portion of the iron is retained in the body discarded. The filtrate is boiled to eliminate HKS, tissues concerned with iron storage. Hahn et al. Fe++ is oxidized to Fe... with H2O,, and Fe (OH)3 (10) gave ferric ammonium citrate to dogs and is preciptated with NH4OH. This precipitate assayed liver and spleen for both ferritin and is dissolved in 1.5 M H2SO4, and 10 mgm. of in- radioactive iron, and found a relatively high level dium is added (radioactive indium is formed from of radioactivity in the ferritin iron. It therefore the cadmium in the silver solder used). Pre- seems highly probable that the radio-iron becomes cipitation of Fe... with cupferron leaves the in- intimately mixed with all of the body iron stores dium behind in the filtrate .(9). The cupferron present when the isotope was injected, and is precipitate is carefully ignited, the Fe2O3 dis- hence available for hematopoietic needs. solved in concentrated HCl, diluted to 3 M, in- It is generally believed that hemoglobin is dium carrier added and the precipitation with "laid down" inside the developing erythrocyte. cupferron repeated. This precipitate is ignited, If some of the iron atoms involved in this syn- dissolved in concentrated HCl, diluted, filtered thesis are radioactive, a proportionate number of (to remove unburned carbon which usually is them will become an integral part of the new present after igniting the cupferron precipitate) hemoglobin molecule within the newly developed and the iron reprecipitated as Fe(OH)3 with cell. When the cell is released into the circula- NH4OH. This preciptate is carefully ignited in tion, its presence in the blood stream can be de- a weighed crucible and the Fe2O3 weighed to detected as long as it remains morphologically in- termine the weight of iron present, tact. When the tagged red cell is destroyed, its RADIOACTIVE ISOTOPES OF IRON IN STUDIES OF RED CELLS 607 hemoglobin derived iron is very rapidly removed PREPARATION OF BLOOD SAMPLES FOR from the plasma, to be reused, to some extent at RADIOACTIVE MEASUREMENT least, in the synthesis of new hemoglobin. Both of the iron isotopes have radiations that In general, each red cell contains approximately are readily absorbed in a thick sample of ma- 0.1 per cent iron, or about a thousand million terial. Thus, for maximum detection efficiency atoms of iron. In the radioactive donors, between all of the iron must be separated from extraneous, 1 and 10 of these iron atoms contains a radioac- material, and the metal plated as a thin uniform of the iron atoms in tive nucleus, the vast majority film on a metal planchet. In this form it can be each red cell being ordinary stable iron. placed close to the window of the counter tube It has been demonstrated experimentally that in the used to measure the activity. This separation is there is no exchange of the radioactive iron accomplished by wet ashing the sample and pre- red cells, either with plasma in vitro or in vivo, cipitating the iron. or with saline in vitro. Thus the method differs The aliquot of donor cells, or the recipient's in important respects from the labelling of red packed cells are transferred, with adequate rins- blood cells by absorbed carbon monoxide (11) ing with water, to a 500 ml. Kjeldahl flask, and or by adsorbed radioactive phosphorus (12). It concentrated H2SO4 added. Three to 5 ml. is furnishes a specific method of detecting the ad- adequate for donor, and 10 ml. for recipient sam- vent of new red cells, and to a certain extent, of ples. Inactive iron is added to bring the total iron following their fate in the circulation, and the content of the sample to 10 mgm. It may be as- disposition made of their contained iron. sumed that 1 ml. of packed red cells contains 1 The normal human erythrocyte is thought to mgm. of Fe. A solution of FeCl3 containing 5 have a life expectancy of about 100 days. Within mgm. Fe per ml. is convenient. The inactive iron this period it is possible to estimate the age of should be added before digestion. the oldest tagged cell in the circulation of a sub- The acid mixture is slowly brought to a boil ject who has been given radio-iron. When these over a bunsen burner. Glass beads aid in re- cells are removed from circulation some part of ducing foaming and bumping. When the water their contained radio-iron will eventually re-enter has boiled off and the acid starts to fume, 1 ml. of the circulation in the hemoglobin of new cells. 60 per cent HClO4 is added, slowly, and this is If this "turnover" be allowed to continue for a repeated a few times until digestion is complete. sufficient period of time, the tagged cells will be Not more than 4 ml. of HCI04 should be added of all ages from birth to death, or representative to the original 10 ml. of acid, and it should never of a "mixed age population" of cells. The ad- be added if the H2S04 has boiled dry, as an ex- ministration of the alternate radioactive isotope plosive compound is formed. As digestion pro- of iron at such a time will result in the production ceeds, the boiling becomes less violent, and more of new cells, the presence of which can be dif- heat can be applied. When digestion is complete ferentiated from the presence of "mixed age cells." the liquid will be a straw yellow color when hot. Thus the behavior of young and old cells may be On cooling it becomes colorless and a chalk white studied in the same subject. precipitate settles out. All organic matter must Hemoglobin solutions prepared from cells ob- be completely oxidized or the subsequent pre- tained from donors prepared with either or both cipitation of iron may be incomplete. isotopes can also be used as radioactive tracer The acid residue is then transferred to a 100 material. ml. centrifuge tube with 1 or 2 washings with 0.1 The utilization of radio-iron, either when given N H2SO4. The iron is then precipitated by the as a ferric salt or as dissolved hemoglobin, or re- addition of concentrated NH4OH to slightly be- sulting from the postinfusion breakdown of tagged yond the neutral point. A fine brown precipitate red cells, as well as the application of the "double which becomes flocculent appears at the end tracer" technique to a wide variety of experimental point. A mechanical stirrer increases the speed and clinical studies will be discussed in subsequent with which ammonia can be added. The tube is communications (13, 14). centrifuged at 1200 r.p.m. for 10 minutes and the 608 PEACOCK, EVANS, IRVINE, GOOD, KIP, WEISS, AND GIBSON supernatant fluid removed by suction. Since the a few drops of 3 M H2SO4 are added to bring it precipitate is fine, it is well to use a filter paper down to about pH 6, and electrolysis is continued over the end of the suction tube. The iron is then for a total of 1.5 hours. A 0.5 ml. sample is with- ready for electrolytic deposition on the copper drawn and tested for completeness of deposition counting planchets. of the iron. If this sample contains more than Iron can be plated quantitatively from an am- 0.0006 mgm. of iron, the pH is readjusted to be- monium oxalate-oxalic acid solution (15). The tween 6 and 7, and the plating continued for an- technique of electroplating described herein is a other half hour. modification of the method of Ross (6). The In testing for iron, advantage is taken of the Fe(OH)3, in the centrifuge tube in which it was intense red color developed when ferrous iron is precipitated, is dissolved in 0.5 ml. of 3 M H2SO4. treated with a, a' dipyridyl in an acid solution. This solution is transferred to the plating cells The sample to be tested is made acid with a drop of (Figure 1), using a saturated ammonium oxalate 3 M H2SO4 and 1 drop each of 1 per cent a, at solution. The final volume is about 25 ml., and dipyridyl in 0.5 M HCl and 10 per cent NaHSO3 the pH of the solution is 4 to 5. is added. Gentle warming develops the color in Electrolysis is carried out at 0.9 amp. with a about 2 mninutes. The amount of iron in solution voltage drop across the cell of about 6 to 8 volts. is determined by comparing the color developed Six cells are run in series to avoid unnecessary with permanent color standards in a simple color power dissipation when 110 v. DC is used as a comparator. These standards are made by visual source of current. A small rheostat and an am- matching of cobalt chloride solutions with the meter serve as a control and a monitor for the color developed in iron solutions of known con- groups. Provision is made for shunting each cell centrations. When properly matched, the stand- through a 6 to 8 volt light bulb of the proper re- ards are sealed into small 100 mm. test tubes and sistance so that any number of the 6 cells can be permanently mounted in the comparator. used at once. At the end of an hour the pH of After the iron has been completely plated from the solution is tested. If it is greater than pH 7, the solution (less than 0.0012 mgm. per ml. re- maining) the planchet is removed and dried with an air blast. It is then coated with a film of light machine oil diluted with benzene 1: 100. BAhEL f CAP The sample is now on a copper planchet 1 inch in diameter and 0.02 inch thick (a form readily available commercially). The iron film forms a 3 SPRims ARE PLACEA circle %8 inch in diameter and contains 10 mgm. F.u,tD1srAAAPR -_ of iron or 2.6 mgm. Fe per cm2. It is now ready A TopAtBorron for measuring.

';f COUNTERS WHICH DIFFERENTIATE BETWEEN THE TWO RADIOACTIVE ISOTOPES OF IRON ON A CLASS ROD BASIS OF THEIR NUCLEAR SPECTRA Deutsch et al (16) in cooperation with the au- DI AT7AAifM JIRC thors made a study of the radiations emitted by Fe59 (47 day half-life). They found that this isotope disintegrates by either of 2 similar pro- 5cKVH"w9R1cf 9ves4- BWS cesses, each with about the same probability of or COPPER PIANCHET occurrence. In one case a beta ray, energetic electron, is emitted with 0.46 million electron volts FIG. 1. SEMI-DIAGRAMMATIC VIEW OF THE APPARATUS maximum followed by a gamma FOR ELECTROPLATING IRON ON PLANCHETS FROM SOLUTION, (MEV) energy, ASSEMBLED ray of 1.10 MEV. In the alternate method of Each unit has individual motor drive. disintegration the beta ray has a maximum energy RADIOACTIVE ISOTOPES OF IRON IN STUDIES OF RED CELLS 609 of 0.26 MEV, and is followed by a gamma ray of a stable isotope, when the nucleus captures an 1.30 MEV. In either case, the beta rays emitted electron from one of the inner orbits of the atom. have an average energy of only about % that of This process, "orbital electron capture," leaves a the maximum. The average energy of all the beta vacancy in one of the inner electron shells which rays is then only about 0.12 MEV. is subsequently filled by an electron from an outer Beta rays are absorbed to about the same ex- shell. As a result, an x-ray characteristic of the tent by equal masses of any material, the fraction Mn55 atom may be emitted. The most abundant transmitted through any in particular being de- and most energetic x-ray emitted has an energy of pendent upon the initial energy of the individual 5.9 kilovolts (Ka, line of Mn). One of these x- beta rays. The counters used for measuring Fe59 rays leaves the atom in about 24 per cent of the are filled with helium to about 70 cm. Hg pressure disintegration.3 Any radiation other than this one and they have a mica window approximately 10 x-ray that might be expected in connection with microns thick. The effect of the counter window thickness on the transmission of these beta rays 3 This x-ray can be emitted only if a K shell electron is shown in Figure 2. This emphasizes the ad- is captured, about 80 per cent of the cases, and in only 30 per cent of these K capture processes is an x-ray emitted vantage of using thin counter windows when meas- from the atom. In the other 76 per cent of the cases, low uring Fe59 (6). energy electrons (Auger electrons) and L and M x-rays Fe55 (5 year half-life) disintegrates into Mn55, are liberated but are not detected.

TNICKEESS or ETAq RAY COvNTER WNDoW&Cad FIG. 2. THE EFFECT OF VARYING THICKNESS OF MICA COUNTER WINDOW ON THE COUNTING RATE OF A SAMPLE OF FE59 The experimental point where no counter window was used was normalized to 100. 610 PEACOCK, EVANS, IRVINE,.GOOD, KIP, WEISS, AND GIBSON

ARGON FiLLD

;J.IJ E

pf.-)yVI, I., 4,':p.-' .: -./.

FIG. 3. THE HELIUM FILLED COUNTER USED TO MEASURE BETA RAYS FROM FE59 IS SHOWN ON THE LEFT The argon filled counter for detection of x-rays from Fe55 is shown on the right. The counters are differential since the x-rays are not absorbed in the helium, and beta rays are absorbed in the beryllium window. this disintegration (e.g.: gamma rays, conversion electrons formed by the x-ray will get to the electrons, etc.) has not been observed with our counting volume. In order for the Ka x-ray of present detection techniques. Table I lists the Mn to cause a count it must be absorbed either in mass absorption coefficients and the thicknesses the gas of the counter, or in the walls of the of various materials necessary to attenuate this counter at a depth not greater than would be 0.0059 MEV x-ray to half value. In every case covered by 14 mgm. per cm 2. the absorption is exponential. The striking thing Thus it is found that a counter with a beryllium that will be noted, besides the relatively small window thick enough to stop 47 day iron beta rays amount of material necessary to absorb the radia- and filled with argon gas at 60 cm. Hg pressure tion, is the side difference in the absorbing power detects about 30 times as many x-rays as the 10 for the same mass of different materials (pro- micron mica window counter filled with helium portional to the cube of the atomic number). at 70 cm. Hg pressure. Thus a 10 micron mica window would transmit These two types of counters (Figure 3) are use- 57 per cent of the x-radiation, and a beryllium ful in double tracer experiments, since each will window 0.76 mm. thick would transmit 55 per detect the presence of one isotope and is almost cent of the x-radiation. Such a beryllium window completely insensitive to the other. would transmit less than 5 per cent of the beta rays Ross (6) has already pointed out that with the from the 47 day iron isotope, while a 10 micron 47 day isotope the fraction of beta rays absorbed mica window will transmit 75 per cent of the beta within a source of the present geometry (2.6 mgm. rays. per cm2) is negligible, but that an appreciable Any agent that will produce ionization in the fraction of the beta rays are absorbed as the source "counting volume" of a Geiger Counter will cause gets thicker (approximately 10 per cent are ab- it to "count." In order for the x-ray to be counted sorbed with a 5.2 mgm. per cm2 source). it must be absorbed in the gas of the counter, or Figure 4 shows the effect on counting rate of near enough to the surface of the walls so that plating increasing amounts of inactive iron with RADIOACTIVE ISOTOPES OF IRON IN STUDIES OF RED CELLS 61t sample containing the same amount of Fe55. The SAMPLE MEASUREMENTS curve is proportional to 1-e Nt where N is The copper planchets on which the iron has Nt been plated are placed on a turntable which auto- the mass absorption coefficient of K Mn x-rays, e maticaly places one sample after the other under is the base of the natural logarithm, and t is the the counters which are to measure the activities of thickness of the iron in grams per cm2. The the iron (Figure 5). Each sample is Kieasured for points represent measured activities using increas- a predetermined time interval (usually 15 or 30 ing amounts of stable iron. There is approxi- minutes). As a new sample comes under the mately 1 per cent self absorption per mgm. of counter its counts begin to register on a counting- sample under these conditions. One readily sees rate-meter. The average counting rate is regis- from this figure that it is even more important to tered in ink on the tape of a recording milliam- keep constant sample weight in the case of Fe55 meter. If both iron isotopes are used in the than with Fe59. experiment the records from a beta ray counter MQ. OF IRON OKAPLANCHEET 7/. iN DIAMETER * 5 10 Is 20._,FK25 a0 I I I III

x~~~

I-. x

x~~~

6L- I I- I L I I I 4 .. I I.uI A z.vA 0.U .Q-A 5.0t 6010i 10- M6. OFIRONPER CM2 1.O FIG. 4. A CONSTANT AMOUNT OF FE55 AND VARYING AMOUNTS OF INACTIVE IRON WERE PLATED TOGETHER The points are proportional to the measured activities and the solid curve is the predicted activity if all detected radiation is the K x-rays associated with this isotope. Both the curve and the experimental points are normalized to 100 at zero thickness of iron. 612 PEACOCK, EVANS, IRVINE, GOOD, KIP, WEISS, AND GIBSON

FIG. 5. APPARATUS FOR MEASURING THE RADIOACTIVITY OF FE55 AND FE59 SAMPLES Two batteries of electroplaters appear in the background. A beta ray counter is mounted on the left and an x-ray counter on the right side of the automatic sample changer, in center fore- ground. Planchets are secured to the plungers. The turntable rotates through 150 every 15 minutes; the counted planchet drops, and the next sample is moved up close to the counter window. Counting rate meters and calibrating apparatus appear at the left, and the pen and ink recorders in the middle background. and an x-ray counter are obtained separately and experiment. Thus Ua is independent of radioac- simultaneously and on separate charts, simply tive decay and counter sensitivity. Ua is propor- by putting both types of counters over the same tional to the relative concentration of radioactive turntable. iron found in the various samples measured. From To insure accuracy in measurement, the ap- the known weight of iron in the aliquot used as paratus is checked regularly for changes in cali- the standard, one can readily calculate the number bration, background, or counter sensitivity. of micrograms of tagged iron corresponding to a Ua of one. METHOD OF REPORTING RESULTS Suppose for example 1 ampul of 5 year iron All activity measurements are reported in terms contained 1 million counts per minute as detected of "unit acti-vity" or Ua. We obtain this quan- at a particular time on an x-ray counter, and that tity as follows: the background counting rate is there was 1 mgm. of iron in the ampul. Three subtracted from each sample reading to give net thousand counts per minute would be a conveni- counts per minute. Net counts per minute in a ent counting rate, so a standard would be pre- sample are divided by the total volume of red cells pared containing 3 micrograms of radioactive iron in the sample and the net counts per minute in the and 10 mgm. of inactive iron. A sample of red radioactive standard used. The radioactive stand- cells having 1500 counts per minute per ml. ard is an aliquot of the radioactive material origi- would have a Ua of 0.50. Assuming all this ac- nally administered to the subjects used in the tivity was from the same iron target as the stand- RADIOACTIVE ISOTOPES OF IRON IN STUDIES OF RED CELLS 613 ard, the blood would contain 1.5 micrograms of that these rays are also absorbed in a comparably the tagged iron from the target per ml. of red cells. short distance in tissue. We can obtain a suffi- It is worth while emphasizing at this point, how- ciently accurate numerical estimate of the maxi- ever, that standards of 5 year iron that are used mum radiation dose by assuming that all the beta continuously for a period of weeks will not have rays and x-rays are completely absorbed within the same counting rate at the end of this time as the blood stream. The actual radiation dose per samples freshly plated. This is due to the absorp- gram will be somewhat less than this estimate, tion of the radiation in a thin oxide film that because of the portion of the rays absorbed by forms over the planchet surface. tissues, especially those surrounding the smaller blood channels. RADIATION RECEIVED BY A DONOR WITH In x-ray practice, tissue radiation doses are ex- RADIOACTIVE BLOOD pressed in roentgen units, and it is convenient to There is one more important consideration in express the radiation doses due to radioactive iso- connection with the physical aspect of this study, topes in these same units. One roentgen is that i.e. radiation that will be received by the donor to quantity of radiation which produces 1 electro- whom the radioactive isotopes are given. static unit (esu) of ions, of both signs, in 1 ml. The activity a donor's blood should have varies (0.001293 gram) of air at 00 C. and 760 mm. Hg with the type of experiment being performed. pressure. Because each ion carries a charge of Let us take as an example a case where about 4.80 x 15-10 esu, 1 roentgen amounts to 1/(4.80 x 3000 counts per minute per ml. of donor red cells 10-1 x 0.001293) = 1.61 x 1012 ion pairs per are necessary in order to obtain sufficient activity gram of air. The average amount of energy re- in the recipient cells with either isotope. quired to produce 1 ion pair in air is (18) 32.5 The present beta ray counters detect approxi- electron volts, or 32.5 x 1.60 x 10-12 = 52 X 10-12 mately 25 per cent of the disintegrations of Fe59, ergs. Because tissue is composed of elements and the x-ray counters .detect approximately 3 per having a low atomic weight similar to air, the cent of the, disintegrations of Fe55. energy required to form an ion pair in tissue may We can assume that essentially all of the gamma be taken as the same as the energy required to ray radiation of Fe59 is absorbed outside the blood form an ion pair in air. Then 1 roentgen repre- stream, and that the large majority of it is not sents an energy expenditure in each gram of tis- stopped inside the body at all. The beta rays of sue of 32.5 x 1.61 x 1012 = 52 x 1012 electron Fe59 have a maximum energy of 0.46 MEV and volts, or of 52 x 1J-12 X 1.61 x 1012 = 84 ergs. an average energy of 0.12 MEV. Hence (17) One ml. of donor blood, at a hematocrit of 40 their maximum range is about 1.5 mm., and their per cent, will then contain 3000 x 0.4 = 1200 average range is about 0.15 mm. of blood or sur- c.p.m. When the efficiency of the counters is con- rounding tissue. The mass absorption coefficients sidered, these counting rates correspond to the (Table I) for the x-rays of Fe55 (5.9 kv.) show disintegration of 1200/0.25 = 4800 atoms of Fe59, or of 1200/0.03 = 40,000 atoms of Fe55. TABLE I 'the maximum energy delivered per gram (0.95 ml.) of blood by Fe59 is thus 4800 x 0.95 X 0.12 Mass absorption Thickness in cm. x 108 = 5.5 x 108 electron volts per minute or Material coefficient in necesry to re- cm2 per gram duce intensity to half 5.5 x 1012 electron volts per week, or 5.5 x He (gas at S.T.P.) 0.5 7800. 1012/52 x 1012 = 0.11 roentgen per week. Be 4.3 0.087 Similarly, the maximum energy delivered per Carbon (graphite) 11. 0.028 Oxygen (gas at S.T.P.) 28. 17.5 gram of blood by Fe55 is 40,000 x 0.95 x 5.9 x Mica* 210. 0.0012 108 = 2.2 x 108 electron volts per minute or 2.2 x Aluminum 130. 0.0019 Iron 100. 0.00087 1012 electron volts per week, or 2.2 x 1012/52 x Argon (gas at S.T.P.) 286. 1.35 1012 = 0.04 roentgen per week. Gold 480. 0.000076 Thus the radiation dose from either isotope in * K20.3 Al203*6 SiO9*2H20. a donor may be taken as 0.1 roentgen per week in 614 PEACOCK, EVANS, IRVINE, GOOD, KIP, WEISS, AND GIBSON the blood stream only, and a negligible dose in the duced separately and obtained in a state of radio- rest of the body tissues. Alternately, if the radi- active purity. These isotopes are prepared in ation is assumed distributed uniformly through all the form of ferric ammonium citrate for intra- the body tissues, then the whole body radiation venous injection into human subjects. The radia- dose ig about 0.007 roentgen per week for Fe59 tion doses in donors and recipients are well below donors, and 0.003 for Fe55 donors. In recipients accepted tolerance levels. the radiation doses are much smaller than in the 2. Blood and tissue samples are prepared for donors, usually by a factor of 50. radioactive measurement by H2SO4- HCIO4 di- The tolerance dose for continuous whole-body gestion and ammoniacal precipitation. The iron in exposure is 0.1 roentgen per day, or about 1 the sample is electroplated from an ammonium roentgen per week to all the tissues of the body oxalate-oxalic acid solution onto a copper disk. (19). 3. Routine measurements are reported in terms It is evident that no radiation effects are to be of "unit activity" which is a ratio of sample to a expected, and none have been observed. We have standard aliquot of the original activity. Correc- built up a total of 48 human donors. Three re- tions for decay during an experiment are thus ceived the isotopes Fel5 and Fe59 mixed (produced eliminated. by bombardment of an iron target), 6 received 4. Geiger-Muller counters are described, which Fe5, and 38 received Fe55, 3 of the latter also re- discriminate between the 2 isotopes of iron when ceiving subsequent ddses of Fe59. The calculated both are present in the same biological sample, dosages in these donors have ranged from 0.05 to thus permitting double-tracer experiments. 0.2 roentgens per week. Three of these subjects have been observed over a period of 4 years, 8 We wish to acknowledge the help we have received for from 12 to 19 months, and 36 for from 2 to 10 from N. J. Grant of the Metallurgy Department, in the months. preparation of the manganese targets; Prof. M. S. Liv- Red cells from these donors have been trans- ingston, Dr. Eric T. Clark and the M. I. T. Cyclotron crew for the many bombardments; Rose Clopman, Martha fused into 160 human recipients in amounts rang- Weeks, Florence Tytell and Eleanor Ryan for preparing ing from 50 to 250 ml. These subjects have been and measuring several thousand samples using this tech- under observation for from 2 months to 4 years. nique. All of these subjects were normal young adult males, with active daily routines. In no instance BIBLIOGRAPHY has any change in blood picture occurred, as evi- 1. Hahn, P. F., Bale, W. F., Lawrence, E. O., and denced by changes in hematocrit or hemoglobin Whipple, G. H., Radioactive iron and its metab- levels (other than could be accounted for by bleed- olism in anemia. J. Exper. Med., 1939, 69, 739. ing), or change in leucocyte count. Regeneration 2. Hahn, P. F., Bale, W. F., Hettig, R. A., Kamen, of red cells following bleeding has apparently been M. D., and Whipple, G. H., Radioactive iron and normal, and several donors have been bled more its excretion in urine, bile, and feces. J. Exper. Med., 1939, 70, 443. than once. Three subjects have married subse- 3. Hahn, P. F., Ross, J. F., Bale, W. F., and Whipple, quent to receiving radio-iron, and have begotten G. H., The utilization of iron and the rapidity of normal children. hemoglobin formation in anemia due to blood loss. In addition, radio-iron, in dosages about equiva- J. Exper. Med., 1940, 71, 731. lent to that received by recipients, has been given 4. Balfour, W. M., Hahn, P. F., Bale, W. F., Pom- to 65 patients on the wards of the Peter Bent merenke, W. T., and Whipple, G. H., Radioactive radiation iron absorption in clinical conditions: Normal, Brigham Hospital, without observed pregnancy, anemia, and hemochromatosis. J. Ex- effects. Many of these patients had anemias, and per. Med., 1942, 76, 15. yet the course of recovery from hemorrhage ap- 5. Hahn, P. F., Bale, W. F., Ross, J. F., Balfour, W. peared to be unaffected by the radiation. M., and Whipple, G. H., Radioactive iron absorption by gastro-intestinal tract. 3. Exper. Med., SUMMARY 1943, 78, 169. 6. Ross, J. F., and Chapin, M. A., The electrolytic sep- 1. Two radioactive isotopes of iron, Fe59 (47 aration of radioactive iron from the blood. Rev. day half life) and Fe55 (5 year half life), are pro- Scient. Instruments, 1942, 13, 77. RADIOACTIVE ISOTOPES OF IRON IN STUDIES OF RED CELLS 615

7. Livingston, M. S., The cyclotron II. J. Appl. Physics, volume by means of two radioactive istopes of 1944, 15, 128. iron. J. Clin. Invest., 1946, 25, 616. 8. Ray, T. W., A rapid means of obtaining manganese- 14. Gibson, J. G., 2nd, Aub, J. C., Evans, R. D., Pea- free iron. J. Lab. and Clin. Med., 1940, 25, 745. cock, W. C., Irvine, J. W., Jr., and Sack, T., The 9. Hildebrand, W. F., and Lundell, G. E. F., Applied measurement of post-transfusion survival of pre- Inorganic Analysis. John Wiley and Sons, New served human erythrocytes by means of two isotopes York, 1929, p. 109. of radioactive iron. In preparation for publication. 10. Hahn, P. F., Granick, S., Bale, W. F., and Michaelis, 15. Classen, A., and von Reis, M. A., Electrolytische L., Ferritin: VI. Conversion of inorganic and Bestimmungen und Trennungen. Berichte, 1881, hemoglobin iron into ferritin iron in the animal 14, 1622. body. Storage function of ferritin iron as shown 16. Deutsch, M., Downing, J. R., Elliott, L. G., Irvine, by radioactive and magnetic measurements. J. J. W., Jr., and Roberts, A., Disintegration schemes Biol. Chem., 1943, 150, 407. of radioactive substances. IV. Fe59. Physiol. 11. Chang, H. C., and Harrop, G. A., Jr., The determina- Rev., 1942, 62, 3. tion of the circulating blood volume with carbon 17. Evans, R. D., Isotopes, radioactive, measurement. monoxide. J. Clin. Invest., 1928, 5, 393. Med. Phys. Yr. Bk. Publishers, 1944, p. 643. 12. Hahn, L., and Hevesy, G., A method of blood volume 18. Gray, L. H., An ionization method for the absolute determination. Acta Med. Scandinav., 1940, 1, 3. measurement of gamma ray energy. Proc. Roy. 13. Gibson, J. G., 2nd, Weiss, S., Evans, R. D., Peacock, Soc. London, 1936, 156A, 578. W. C., Irvine, J. W., Jr., Good, W., and Kip, 19. Radium protection. National Bureau of Standards A. F., The measurement of circulating red cell Handbook No. 23, 1938.