U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content













Chapter XII



The Bovine and Human Albumin Programs

Part I. Bovine Albumin

In the spring of 1940, when medical resources first began to be mobilized with the realization that the United States would eventually enter World War II, the use of blood serum for shock and other conditions was limited to a few pioneer workers in a few medical centers. Serum albumin, which was to prove the mainstay of the Navy, had not yet been developed. Work with bovine albumin was limited to a few pilot studies, chiefly by Wangensteen at the University of Minnesota (1).

The bovine albumin program in World War II began in late 1940 and progressed in a series of highly encouraging developments until July 1942, when the first real setback was encountered, in the form of a fatal case of apparent serum sickness. Because the potentialities of this form of therapy were believed to outweigh the risks, the program was continued cautiously. A number of serious reactions, however, occurred in volunteers used for testing purposes, and, when a second fatal case of serum sickness of an unusual type was encountered in February 1943, further work with bovine albumin was regarded as unjustified, and the program was officially discontinued the following month. It was a truly discouraging end to a highly promising project.


One of the questions raised at the first meeting of the Committee on Transfusions, Division of Medical Sciences, 31 May 1940 (2), was the possible development of a substitute, preferably synthetic, for human plasma. At this same meeting, "in the interest of clear thinking," it was proposed that protein chemists be brought into the work, and the assistance of Dr. Edwin J. Cohn (fig. 73), Department of Physical Chemistry, Harvard Medical School, was obtained (p. 336). A synthetic substitute was not developed during World War II, but, as just indicated, a great deal of time and effort went into the development of a bovine substitute for human serum albumin.

The first step in the program was a report, at the first meeting of the Subcommittee on Blood Substitutes on 30 November 1940 (3), of previous work with bovine and human plasma by Dr. Owen H. Wangensteen, Department of Surgery, University of Minnesota Medical School. His presentation, like his first publication on the subject (1), made it clear that the intravenous clinical administration of bovine plasma had not yet been established as a safe routine hospital procedure. His work, however, had indicated that this agent could be


FIGURE 73.-Edwin J. Cohn, Ph. D.

given by this route to some patients in fairly large quantities. It was Dr. Wangensteen's opinion that, when the possibilities and limitations of this method became clear, it was likely that it would become a practical hospital procedure "useful in civil as well as in war surgery for various purposes having to do with contracted blood volumes and protein stores."

The first decision in the bovine albumin program concerned the agent to be used; namely, serum, plasma, or a purified fraction of one or the other. As the experiments in the Harvard laboratory progressed (4), it became clear that the albumin fraction had many desirable physiologic and chemical properties and that it was more stable, less viscous, and less antigenic than whole plasma. It also had a larger relative osmotic effect. Chemically, Dr. Cohn reported, there appeared to be no difference between human and bovine albumin. They were the same as to solubility, isoelectric point, electrical charge, mobility, electrophoretic pattern, sedimentation constant, and shape. The difference between them could be detected only by precipitin tests.

Preparation - Bovine serum albumin was first prepared by a large-scale ethanol-water fractionation method, with purification by isoelectric precipitation. The chief advantage of this technique, which had been developed in the Harvard laboratory, was that at the end of the process, the material could be passed through a Seitz filter. The globulin content was undesirably high, 1 percent, and the single commercial firm (Armour Laboratories Division of Armour & Co.) attempting to produce bovine albumin was having a great deal


of difficulty in reducing it. This firm, incidentally, clearly understood that if an acceptable product was finally accomplished, it would have no monopoly on the process.

Beginning with the meeting of the Subcommittee on Blood Substitutes on 19 April 1941 (4), successive reports were made on the clinical testing of bovine albumin. It was Dr. Cohn's opinion then, and continued to be his opinion until almost the end of the project, that reactions observed were more probably the result of globulin, which might be present in such small amounts as not to be demonstrable by present techniques, than the result of albumin per se.

Dr. Cohn listed as immediate objectives in clinical tests:

1. Further proof that beef albumin would replace blood lost in acute hemorrhage.

2. Proof that it would replace plasma in shock. At this time, Dr. Cohn could not recommend it for this purpose.

3. Statistical studies on such matters as the incidence of serum disease and the safety of multiple injections.

It was the opinion of the group at Harvard, based on their immunologic studies, that beef albumin was remarkably inert in the circulation, in which it remained detectable for long periods of time, and was apparently a fairly poor antigen in the human bloodstream.


Since bovine albumin does not appear in human urine, animal experiments were necessary to determine its ultimate fate in the body. Studies conducted by Dr. Orville T. Bailey, Harvard Medical School, on rabbits indicated that such tissue changes as occurred were apparently reversible and were of biologic rather than clinical interest. Nothing resembling amyloid was observed(5, 6).

By July 1941, the crystallized bovine albumin originally produced had been greatly improved in purity. The coloration always present in all albumin preparations disappeared upon recrystallization, a phenomenon which suggested that, in both human and bovine products, the coloration was chiefly dependent upon concentrations of globulins, especially beta globulin, in them.

By April 1942, progress had been so satisfactory in all respects that a detailed report on the crystallization of bovine albumin was made to NRC (National Research Council), acting for the Committee on Medical Research, Office of Scientific Research and Development (7). This report included instructions for the preparation of crystallized bovine albumin, a tabulation of its physical constants in solution, a complete report of its molecular properties in comparison with those of human serum albumin, a report on the experimental histologic effects of the crystalline preparation, and the course and progress of commercial preparation of the material at the Armour Laboratories. A progress report on the clinical experience to date was also included.

When no detectable globulin was found in the available preparations by chemical tests, Dr. Charles A. Janeway undertook a study of bovine albumin


by immunologic methods. His results led him to conclude, for two reasons, that he was measuring a residual impurity:

1. The impurity diminished with each successive recrystallization. This was strikingly illustrated in one preparation, which was recrystallized four times, each time with a decrease in the precipitable substance.

2. The same results were always obtained, no matter what preparation of crystallized bovine albumin was used in adsorbing the antiserum.

The practical application of these results, in Dr. Janeway's opinion, involved the solution of two problems:

1. The development of a more potent antiserum to provide for more readily detectable precipitation. It was thought that an alum-precipitated antigen might be useful.

2. The development of a more accurate method of quantitating the amount of precipitation. It was thought that nephelometry would be simpler than the Kjeldahl nitrogen technique then in use.

Plans were made for further testing by immunologic methods.

At the conference on 16 July 1942 (8), plans were made to study a number of points concerning bovine albumin, but it was agreed that, important as were these matters, none of them should be permitted to divert attention from securing answers to two questions of primary importance to the Armed Forces, (1) the safety of repeated large clinical doses of bovine albumin and (2) its effectiveness in shock.

At this conference, the first instance of serious serum sickness was reported (p. 330), following the use of material that had been recrystallized four times. In discussing the case, Dr. Cohn said that when bovine albumin of satisfactory stability had been obtained, the globulin component had been reduced to less than 2 percent, but preliminary work had shown that no such amount of globulin could be permitted. Crystallization was therefore undertaken, and clinical tests were begun with the new product in November 1941.

Since serum sickness had not been eliminated by the use of even recrystallized bovine albumin, material was being sent to a number of investigators who thought that they could despeciate the molecule, and similar studies were being conducted in Dr. Cohn's laboratory.1

Three months later (in October 1942), Armour Laboratories was producing bovine albumin with 0.01 percent globulin, and the Harvard laboratory was making a product with 0.001 percent (9).


April 1941-June 1942

The development of the bovine albumin program in respect to clinical testing was reported at various meetings of the Subcommittee on Blood Substitutes and at various conferences, and is most conveniently discussed chrono-

1The clinicians testing bovine albumin reported to Dr. Janeway, who summarized their data and passed it on to Dr. Cohn for further evaluation.


logically. Dr. Wangensteen, who was continuing his personal studies with this agent (10), also reported at each meeting.

Encouraging reports were made at the meetings on 19 April 1941 (table 9) (4); 8 May 1941 (11); 10 March 1942 (12); 12 May 1942 (13); and 23 June 1942 (14). At the May 1942 meeting, the results seemed so promising that it was agreed to expand the testing program, heretofore confined to Peter Bent Brigham Hospital and Dr. Wangensteen's clinic, to certain other selected hospitals as supplies permitted.2

TABLE 9.-Results of clinical testing with bovine albumin to 19 April 1941


Persons injected









Armstrong No. 3





Armour's IV W
     15% ethanol





Armour's IV W
     40% ethanol





Armour's VII
     40% ethanol




??Slight, 2 hr. later.

Armour's VII
     40% ethanol





Armour's VII
     40% ethanol




Serum sickness after 10 days.

Armour's VII
     40% ethanol




??Slight, immediate, when
injection too rapid.

Armour's VII W
     rework brown oil:





(1) 40% ethanol





(2) 10% ethanol





July 1942

7 July.-At a special meeting of the Committee on Medical Research on 7 July 1942 (15), there was a full discussion of the criteria to be employed in the recommendation of bovine albumin for military use. Capt. C. S. Stephenson, MC, USN, thought that 50-60 persons should first be injected with doses of shock size, repeated within 1 or 2 weeks, and followed by reinjections of 25 to 50 gm. after the blood was free of circulating antigen. Dr. Robert F. Loeb deplored the selection of any special number of test subjects. Dr. Wangensteen thought it would be safe to recommend bovine albumin after it had been used successfully in all conditions for which plasma was used. The meeting was

2Through error, this promising news was released to the press, and it had an immediately adverse effect on Red Cross solicitation of blood donors.


reminded by the representative of The Surgeon General that no blood substitute would be practical under combat conditions if a skin test was necessary before it was used. No action was taken.

16 July -At the Conference on Bovine Albumin on 16 July 1942 (8), the outlook, with one exception, seemed very encouraging. Only four reactions had been encountered in 86 injections of crystalline bovine albumin, in amounts ranging from 10 to 50 gm., against the same number in 35 injections of amorphous albumin. Three of the four reactions with the crystalline substance were pyrogenic, all from the same preparation and to be explained by allowing the solution to stand too long at room temperature between solution and sterilization.

The fourth reaction introduced a disturbing note, for it was a case of possible serum sickness. The summarized case history follows:

Case 1 -A 62-year-old Italian, casually selected for testing during an uneventful convalescence from herniorrhaphy and orchiectomy (and with an overlooked previous history of a 20-pound weight loss and mild epigastric pain during the preceding 6 months), was given 12.5 gm. of crystallized bovine albumin twice, at a 6-day interval. The first injection was with a preparation suspected, because of results of the thermal rabbit test, of being pyrogenic. On the 10th day after this injection (the 4th day after the second injection), he developed fever, anemia, arthralgia, edema, purpura due to capillary fragility, urticaria, hypoproteinemia, and nitrogen retention. The fact that bovine albumin disappeared more rapidly than usual from his bloodstream, and that he had a strongly positive skin test as soon as it disappeared, favored the diagnosis of serum sickness. On the other hand, three of seven blood cultures were positive for Bacillus pyocyaneus. Contamination was suspected, and it is impossible to avoid the suspicion that this patient had a primary bloodstream infection.

In the discussion that followed the presentation of this case report, the Navy representative stated that he was not too much disturbed by a single instance of serum sickness, especially in a patient with so many other complications. In Dr. Loeb's opinion, in which he was joined by others, unless testing was proceeded with more rapidly, and with the use of larger doses, the problem of bovine albumin, from the military point of view, would become academic.

There was a considerable discussion at this conference of the risks run by subjects and investigators in this kind of work. Dr. Wangensteen thought that the administration of bovine albumin was a justifiable therapeutic procedure, which carried no greater risks than transfusion. Others took the position that reinjections into healthy subjects carried definite risks and must usually be on a voluntary basis. It was the opinion of the meeting that signed releases were not binding and simply emphasized the risks. Dr. Alfred Blalock thought that medical students should no longer be used as subjects.

The conference ended, the minutes relate, "on an optimistic note." All present believed that any risks involved in clinical testing were justified by the encouraging results obtained to date and by the urgency of the needs of the Armed Forces.

At this conference, Dr. Wangensteen reported that 40 to 60 percent of 120 of his patients injected with whole bovine albumin developed serum sickness,


usually about 4 days after the injection had been concluded. He believed the case reported at this meeting to be an instance of serum sickness, though modified by some other factor. He had had no reactions in about 60 patients treated with the amorphous and crystalline preparations of serum albumin prepared by Dr. Cohn and his group, which had been injected in amounts of 25 to 50 gm. in 200 cc. of physiologic salt solution. Six of these patients had previously reacted to injections of bovine plasma.

October 1942

When a Conference on Albumin Testing was convened on 19 October 1942 (16), part of the picture was extremely encouraging. There had been only one immediate anaphylactoid reaction in 170 first injections, and none in over 100 reinjections. Of 36 injections made with a purified, globulin-free, well-tested bovine albumin, 25, in which no reactions had been experienced, were second, third, or fourth injections.

The rest of the picture, Dr. Janeway reported, was highly discouraging. The patient with serum sickness reported on at the 7 July meeting still had severe renal damage. There had been a high incidence of serum disease in the volunteers being studied at Welfare Island, the Norfolk Prison Colony, and the Massachusetts Department of Correction. Although the testing group had been directed to proceed with all possible speed and to inject as many persons as possible three times each over a 3-month period, Dr. Janeway considered that the risks had become too great, and, on 18 September, he had telegraphed the investigators to discontinue all clinical testing until further notice.

Dr. Wangensteen expressed himself as having been surprised and disappointed when he had been asked to discontinue testing. He regarded his results with the new preparation as most encouraging. He had given 126 injections to 80 patients with eight immediate reactions (one anaphylactoid) and four delayed reactions. One of the latter occurred 21 days after the injection and was an instance of severe serum sickness; the situation was confused by the fact that the patient had also had tetanus antitoxin and that her skin test, which was positive to horse serum, was negative to bovine albumin during the latter part of her illness.

Two other reports were made:

1. Dr. Blalock, Johns Hopkins Hospital, reported the treatment of four patients in traumatic shock, with good clinical responses to 25-50 gm. of bovine albumin. One patient, with multiple fractures, had died of fat embolism. At the end of a month, the other three had negative skin tests, but two showed leukopenia.

2. Dr. James T. Heyl reported on his work in Boston. Up to this time, there had been no significant reactions in 25 patients and medical students in whom the emphasis had been on reinjections. When 200 prisoner volunteers were obtained, the risk of serum sickness was one case in 180 injections and


reinjections. There were, however, 21 delayed reactions in 66 injections in this group, 1 of which was fatal and 20 of which were severe enough to require hospitalization.3 A followup revealed three additional delayed reactions.

Case 2 -In the case which ended fatally, the presenting symptom, which was also the most prominent symptom, was acute arthralgia, which appeared in the hip 19 days after the injection. Fever, myalgia, and malaise were moderate. Chills and visceral pain were not present. The patient died in his sleep suddenly and unpredictably on the eighth day after symptoms appeared, 2 hours after he had waked spontaneously for 20 minutes. When his bed was changed at this time, he seemed perfectly normal.

Autopsy, performed by Dr. Bailey, revealed significant findings only in the heart and lungs. The heart, which weighed 400 gm., was flabby, and the myocardium was soft. Histologic examination showed extensive interstitial edema of the myocardium, with infiltration by mononuclear and polymorphonuclear leukocytes. In focal areas, especially in the interventricular septum, coagulation necrosis of myocardial fibers was present, in association with more extensive cellular infiltration.

The right lung weighed 790 gm. and the left, 730 gm. There was extensive pulmonary edema.

The combined changes in the heart and lung were considered to be the cause of death, but the pathologist stated that, if he had performed the autopsy in ignorance of the situation, he could not have determined that a foreign protein had been injected.

The kind of delayed reaction observed in these recent cases resembled the syndrome in the case of serum sickness that had occurred earlier (p. 330). It occurred, on the average, about 14 days after the injection, which most often was the first. It did not resemble the type of reaction previously observed after injections of amorphous bovine albumin.

A vigorous attempt to determine the component of the albumin preparations responsible for the recent reactions was already underway. Dr. Cohn was sure that stable crystallized bovine albumin could be prepared by careful control of the conditions of laboratory testing. It was essential to have tests that could pick up alterations in the product. Until some satisfactory animal test was developed, nephelometry could be used to detect the lack of heat stability characteristic of preparations containing denatured albumin.

In spite of the discouraging data presented at this meeting, it was the consensus that the military situation justified the continuation of both experimental studies and clinical testing. Dr. Cohn and the group working with him were authorized to take all the time necessary to develop tests to distinguish between good and bad preparations of bovine albumin. When they considered it safe, clinical testing could be cautiously resumed.

When the Subcommittee on Blood Substitutes met the day after the Conference on Albumin Testing (9), it accepted this recommendation, and the proposed specifications for further testing were outlined in an addendum to the minutes. It is significant and prophetic, however, that Lt. Col. (later Col.) Douglas B. Kendrick, MC, wrote in the margin of his copy of the conference minutes, "Bovine albumin is now dead as a dodo."

3Dr. Heyl and others connected with the project paid tribute to the courage and unselfishness of the volunteers from the Commonwealth of Massachusetts Department of Correction and the Norfolk State Prison Colony. They clearly understood the risk involved before they volunteered, and their cooperation continued in spite of the high incidence of reactions and the fatality that followed one of them.


December 1942

At a meeting of the Subcommittee on Blood Substitutes on 15 December 1942 (17), Dr. Cohn reported a marked improvement (10 to 20 times) in the stability of bovine albumin preparations. Immunologic studies, however, had still failed to develop satisfactory tests to distinguish lots of albumin that had produced clinical reactions from those that had not. Dr. Wangensteen was extremely desirous of resuming clinical testing, on the basis of his previous good results, and its cautious resumption was authorized, on the ground that further criteria of safety could be developed only by such observations.


The official bovine albumin program ended at the meeting of the Albumin Testing Group on 22 March 1943 (18), with the report of a serious reaction in the first patient who had been injected with the new preparation of bovine albumin. The stability of this product was 20 times that of the preparations which had caused the previous serious reactions. The globulin content was estimated at 0.008 percent.

Case 3.-The patient, a 55-year-old white hemiplegic, had had a practically stationary course for the last 3-4 years. He was aphasic, and his spinal fluid Wassermann reaction was positive. He was injected with 25 gm. of the new bovine serum albumin on 20 February 1943. Fifteen days later, without previous symptoms, he suddenly developed a purpuric rash on both legs. He had a chill 3 days later followed by 11 days of fever, during which there were two other chills. He had migratory pains in the limbs, chest, precordium; urticarial as well as purpuric lesions over the legs and lower trunk; and general malaise. For the 10 days before the 22 March meeting, he had felt well.

Laboratory tests showed an increased sedimentation rate; decreased red blood cell counts; decreased hemoglobin values; signs of impaired liver function; a sharp decrease in serum albumin, with no change in the serum globulin; a significant decrease in excretion by the phenolsulfonphthalein test; and no change in the nonprotein nitrogen of the blood.

It was generally agreed that this reaction, whatever its cause, was entirely similar to those previously observed. As soon as Dr. Cohn had heard of it, he had written to Dr. Alfred N. Richards, chairman of the Committee on Medical Research, NRC, informing him of the reaction to a preparation which, like another prepared at the same time, he described as "beautiful." Dr. Wangensteen and Dr. Blalock had used the same preparation without any reaction, immediate or delayed, and a patient on whom Dr. Heyl had used it was asymptomatic at the end of 3 weeks.

In view of the reaction which had occurred with such a greatly improved product, Dr. Cohn had no choice but to consider that his earlier working hypothesis, that the first reactions with bovine albumin were due to the instability of the product, was no longer tenable. He also believed that he had no choice but to recommend that all further chemical and clinical efforts to develop bovine albumin as a project by the Committee on Medical Research


should be discontinued. The Albumin Testing Group approved his action and recommendation, although Dr. Wangensteen continued to be puzzled by the difference between his own results and those of others. To date, he had injected 90 patients with crystalline bovine albumin with only one mild immediate anaphylactoid reaction and only eight delayed reactions, seven of which were mild.

Dr. Cohn regretted that the early promise of bovine albumin had not been borne out. Regardless of the success or failure of future attempts to develop it, he pointed out that if Dr. Walter B. Cannon had not suggested the original plan of developing a blood substitute from animal sources (p. 76), the present method of plasma fractionation, which was yielding human serum albumin, antibody globulins, fibrinogen, and thrombin, might not have been devised.

The action of the Albumin Testing Group was approved at the 23 March 1943 meeting of the Committee on Medical Research, Office of Scientific Research and Development (19), and steps were taken to dissolve the contracts in existence with the various testing groups.

Dr. Cohn emphasized that 177 patients had received injections of crystalline bovine albumin before a severe reaction was encountered. This fact was a challenge which called for an explanation, but solution would be a long term project. Although the hope was expressed that independent efforts would be made to continue testing of bovine albumin, it was realized that this would not be practical without the official sanction of the Committee on Medical Research, because of the risks to patients.


Although all intensive work on bovine albumin ended in March 1943, certain other developments might be mentioned to complete the record.

Antimicrobial agents - All the reactions caused by bovine albumin had followed injections of preparations containing Merthiolate (thimerosal), and Dr. Cohn thought that its presence might play some part in them (20). Clinical results were not conclusive, and animal experiments were undertaken to determine the possible influence of Merthiolate upon antigenicity, especially upon the Arthus phenomenon. Early experiments were suggestive, and it was decided for the future to release bovine albumin without Merthiolate for all clinical testing.

Despeciation - All work on the despeciation of bovine albumin had been extremely disappointing. In some instances, it had not been accomplished at all, and, in others, extremely toxic substances had been produced.4

Clinical testing -The final report of Dr. Wangensteen's contract with the Office of Scientific Research and Development, dated 10 September 1943 (22),

4Nine months after the U.S. program had been abandoned, a group of English workers reported the successful injection of despeciated bovine albumin (21).


covered 139 patients and included a series of 25 subjects injected with the same preparation from the Harvard laboratory with no reactions. His experience indicated to him that bovine albumin would be a satisfactory blood substitute.

At the 2 June 1944 meeting of the Subcommittee on Blood Substitutes (23), Dr. Wangensteen reported on 15 patients who had been injected with the latest preparation of bovine albumin prepared in the Harvard laboratory and all of whom had received injections from previous lots. Two patients who had had severe reactions on their initial injections had mild purpuric reactions after the second injections. There were no other reactions in this series.

In April 1944, Dr. Wangensteen (7) reported the injection of 83 additional patients with solutions of bovine albumin, with three pyrogenic reactions, all susceptible of explanation by technical errors in preparation of the equipment. There were eight delayed reactions, with incubation periods ranging from 14 to 24 days, all of the type seen in mild serum sickness. The Rumpel-Leede test was positive, but there was no purpura. The incidence of serum disease in his experience thus remained at about 10 percent, as in his original report.

A comparative study on human and bovine albumin by Heyl, Gibson, and Janeway (24), published in November 1943, showed no essential difference in the ability of these agents to draw fluid into the circulation after acute blood loss. No significant immediate or delayed harmful effects were observed during the 3-month period the subjects of the test were followed, but the investigators would not commit themselves, without more extensive clinical tests, as to the safety of using a protein of animal origin in man.

Offers and suggestions.-During the war, occasional letters were received from lay persons suggesting that the blood of beef cattle be used in the treatment of wounded men. It was pointed out in the replies that an extensive experience had revealed no way of making animal blood safe for human use.

On occasion, suggestions of the same kind were made from foreign countries, some purporting to have solved all the problems connected with the use of bovine plasma. These suggestions were similarly answered.

No serious attempt was made to revive bovine albumin as a substitute for human serum albumin after the Office of Scientific Research and Development abandoned the idea in 1943.

Part II. Human Serum Albumin


The first significant work on blood serum was done in 1918 by Mann (25), who observed that the parenteral injection of homologous serum in surgical shock, particularly when large doses were used, produced results as good as, or better than, were obtained by any other method. He therefore suggested that this agent might be of value in conditions in which it could be stored and whole blood could not be obtained.


When Dr. Max M. Strumia and his group at the Bryn Mawr Hospital (26) began to use fresh and preserved serum in the treatment of severe infections, they also encountered frequent severe reactions, even when the serum was homologous and caused no agglutination of the recipient erythrocytes. In 1936, Elliott (27) proposed that untyped serum and plasma be used in obstetric shock (p. 266), his reasoning being that the maintenance of osmotic pressure is a function of the plasma proteins and that the need for replacing lost blood volume is more important than the need for replacing red blood cells. Meantime, Stokes, Mudd, Flosdorf, and their associates at the University of Pennsylvania (28-30) were using lyophilized human serum5 prophylactically and therapeutically in various infectious diseases.

In 1937, Fantus (31) advocated serum in burns and in shock without hemorrhage, because of its therapeutic and natural immunizing properties and also because these conditions were usually associated with an excess of red blood cells. In 1940, Strumia, Wagner and Monaghan (26) reported their results with fresh and preserved plasma in a variety of diseases, but found themselves handicapped by frequent, severe reactions, which they attributed to the changes induced in the serum by the lyophile process. In 1938, after a considerable experimental experience, Bond and Wright (32) suggested the use of regenerated lyophilized serum in hemorrhage and traumatic shock. Mahoney (33) reported similar experimental results the same year.

It is typical of the status of human serum albumin at the beginning of U.S. participation in World War II that, at the organization meeting of the Committee on Transfusions on 31 May 1940 (2), only whole blood and dried and liquid plasma were discussed. Serum albumin was not mentioned at all, though the hope was expressed that possible substitutes for human plasma would be found, especially a substitute that could be prepared by synthesis.


The development of serum albumin in the treatment of shock really began at the first meeting of the Subcommittee on Blood Substitutes on 30 November 1940 (3), with the suggestion, already mentioned, of the chairman, Dr. Cannon, that "it would be in the interest of clear thinking" if protein chemists were brought into the picture (p. 325). In response to the suggestion, Dr. Cohn, whose work on plasma fractionation was already outstanding, undertook his work on serum albumin. Albumin was selected rather than any other component of plasma for several reasons: that it constitutes 65 percent of the proteins of plasma, that it exerts 85 percent of the osmotic pressure of

5Reichel was the first to apply the term "lyophilization" (copyrighted by Sharp & Dohme) to the process by which serum was dehydrated (30). The root word, which means solvent-loving, was convenient and well chosen because it emphasized the noteworthy characteristic of the product, its remarkable solubility, which is a result of both the unaltered lyophilic properties of the serum proteins and the physical structure of the porous solid. The antibodies and complement of the serum suffered no detectable loss in processing, and the rate of subsequent deterioration was reduced to a small fraction of the losses in the liquid state.


plasma, and that it was hoped that it could be packaged in small kits and used in preference to plasma whenever space limitations indicated.

At the 19 April 1941 meeting of the subcommittee (4), after a discussion of the best methods of collecting and dispensing both plasma and serum, the following recommendation was adopted: "* * * the consensus of the committee is that either serum or plasma reduced to either a frozen or a dried state is acceptable and that production should proceed at once with the understanding that in time other recommendations may be made."

The Red Cross had already begun to supply Dr. Cohn with 15 bloods per week, and, at this meeting, he presented three exhibits of albumin prepared by plasma fractionation. The first was a tube containing 10 gm. of serum albumin in 11 cc. of water. The solution appeared slightly discolored because it contained traces of globulin, but it was described as a very stable liquid. It had a high viscosity, like that of heavy machine oil. The second tube contained albumin in a 25-percent solution; this was a clear monogenous fluid. The third tube contained 10 gm. of crystalline albumin. Dr. Milton V. Veldee preferred the 25-percent solution but Dr. Cohn thought the crystalline preparation would be more satisfactory for shipment or for storage in bulk.


At the meeting of the Subcommittee for the Standardization of Dispensing Equipment (11), a number of experimental studies were reported, and Captain Kendrick reported the first clinical use of human albumin in traumatic shock:

Case 4.-A 20-year-old man was admitted to Walter Reed General Hospital, Washington, D.C., in May 1941, 16 hours after he had sustained bilateral compound comminuted fractures of the tibia and fibula, fractures of five ribs; and associated pleural damage, pneumothorax, and subcutaneous emphysema. He was confused and irrational, with a blood pressure of 76/30 mm. Hg. After he had been given two units of albumin (each approximately 25 gm.), over a 30-minute period, the pressure rose to 106/70 mm. Hg, and two hours later, after insertion of a Kirschner wire, reduction of one of the fractures, and application of a cast, it was 130/80 mm. Hg. Over the next 12 hours, the patient received 1,250 cc. of fluid by mouth and 1,000 cc. of physiologic salt solution subcutaneously. The systolic pressure remained above 130 mm. Hg during this period, with occasional elevations to 150 mm. Hg. There was no evidence of circulatory failure at any time after the administration of the albumin.

The red blood cell count on admission was 4.1 million per cu. mm., the hemoglobin 14.5 gm. percent, and the hematocrit 44. Twelve hours after the administration of albumin, the red cell count was 3,690,000. Forty-eight hours later, it was 3,780,000 per cu. mm., and the hematocrit was 35 percent. The hemoglobin level was between 12 and 12.5 gm. percent. There were no urinary abnormalities at any time.

At the meeting of the Committee on Blood Transfusions and the Subcommittee on Blood Substitutes on 23 May 1941 (5), it was recommended that the albumin produced in Dr. Cohn's laboratory be tested as rapidly and as extensively as possible. If the clinical tests proved satisfactory, Col. (later Brig. Gen.) Charles C. Hillman, MC, for the Army and Captain Stephenson for the Navy stated that serum albumin would be accepted by the services.


Production of serum albumin in the Harvard laboratory rose to 500-800 gm. per week in July (6). At the 19 September 1941 meeting of the subcommittee (34), Dr. Cohn announced that 3,407 gm. had been prepared or was in preparation; that 447 grams had been used clinically on 11 patients, one of whom had been reinjected after 13 days and another after 2 months; that the dosages had ranged from 7.6 gm. to 61.2 gm.; and that there had been no reactions of any kind.

Further reports were equally good. Up to 31 October 1941 (35), 50 patients had been injected with 788.4 gm. of albumin with no reactions. By 11 February 1942 (36), 125 patients had been injected with 4,247 gm., with only four mild reactions and one moderately severe reaction. The latter followed the use of albumin prepared from dried plasma secured from broken containers. When it was used, it was noted that it was of darker than usual coloration. This reaction occurred after 60 gm. of the albumin had been given to a patient who had lost 1,080 cc. of blood.

The Pearl Harbor experience - At the Conference on Albumin on 5 January 1942 (37), Dr. (later Brig. Gen.) Isidor S. Ravdin, who had just returned from Hawaii, reported the administration of albumin to seven very severely burned patients injured 10 days earlier at Pearl Harbor. All were edematous and all were losing plasma at the site of their burns. Some of them were so edematous, in fact, that albumin had to be injected into the femoral vein because other veins could not be located. The Naval Hospital had had only 40 units of dried plasma for all its casualties, and some of these patients, in the emergency, had received too much salt solution.

All seven patients were given albumin, and all showed prompt clinical improvement, including one whose state was so critical that the administration of albumin to him was debatable. There was no question as to his response: He was unconscious in the morning when he was given 250 gm. of albumin. In the afternoon, he was talking but was disoriented. The following morning, he was given the same amount of albumin. Twenty-four hours later, the edema had disappeared and he was taking food by mouth.

Most of these patients had hypoproteinemia and had very low hemoglobin readings and red blood cell counts when they were first seen. Some had as little as 8 gm. percent of hemoglobin, 3.5 mg. percent of protein, and 2½ million red blood cells per cc. After the second injection of albumin, all showed hemodilution. Urinary outputs were not recorded, since most of the patients were incontinent, but the rapid disappearance of edema in all cases suggested the excretion of large quantities of urine.

The single reaction occurred in a patient who had had reactions after each earlier injection of plasma. He had a chill after the first injection of albumin but no reaction after the second.

The injections had been accomplished with some difficulty because of the poor quality of the gum-rubber tubing and the small diameter of the rubber latex tubing which had been sent to Hawaii with the albumin because of the emergency.


From his experience at Pearl Harbor, Dr. Ravdin concluded:

1. Albumin accomplished osmotically everything that normal human plasma could accomplish.

2. It produced hemodilution

3. Laboratory studies showed that it caused a rise in the albumin fraction of the blood and often in the globulin fraction as well. The fibrinogen component was not investigated.

4. Albumin was a very satisfactory agent in patients who needed protein and who had not had too great a red blood cell loss.

The formal testing program continued into early 1943. By the end of February of that year (38), the total number of injections had reached 550, of which 72 were for hemorrhage and for surgical or posttraumatic shock and 48 for burns. Reactions continued insignificant and mild.

Testing Procedures

The following criteria were established after clinical testing had been in progress long enough to indicate what direction they should take (35):

1. Serum albumin should be given to patients in shock, who, if possible, should receive only albumin. Normal subjects should be tested for possible reactions. In all cases, attention should be paid to whether or not fluids were withdrawn from the tissues.

2. The initial dose should usually be 100 cc. of 25-percent solution (1 unit), which is roughly the equivalent of 500 cc. of citrated plasma (39). Administration should take about 20 minutes. The original injection should be repeated in 15 to 30 minutes if the desired response is not obtained. Not more than 10 units (250 gm.) of serum albumin should be given to any patient in 48 hours.

3. Vital signs should be recorded at specified intervals: In experimental subjects, at 30-minute intervals for 2 hours and then at 4-hour intervals for 24 hours; in shocked subjects, at 30-minute intervals as long as shock persists, then at 4-hour intervals for another 24 hours. Hematocrit determinations should also be made before the albumin is administered and at regular intervals thereafter in relation to the patient's progress.

4. Concealed hemorrhage is a possibility to be borne in mind, for it may occur or recur as the blood pressure returns to normal. This risk was evident in a patient at Grady Hospital, with a stab wound of the chest, who responded well to albumin but collapsed suddenly when the blood pressure had reached normal, from hemorrhage from a severed mammary artery undetected until albumin had become effective.

5. In extremely dehydrated patients, additional fluid and electrolytes must be given, since albumin draws fluid into the blood at the expense of other body tissues.

6. Serum albumin is not a satisfactory agent in severe anemia, since it simply increases the circulating blood volume without adding red blood cells.


The first formal action on the use of serum albumin in the Armed Forces was taken at the Conference on Albumin on 5 January 1942 (37), when it was recommended to the Surgeons General of the Army and the Navy that, in


addition to continuation of the plasma program, serum albumin be immediately adopted for clinical use for the following reasons:

1. Albumin can be packaged and stored in less than a tenth of the space required for the standard Army-Navy package of dried plasma.

2. It is ready for immediate emergency use, without regeneration.

3. It is stable in solution in temperatures as high as 113° F. (45° C.) for protracted periods.

4. Its adoption will accelerate and supplement the procurement of satisfactory blood substitutes for military use.

It was further recommended at this conference:

1. That the Surgeons General request the Red Cross to secure voluntary blood donors for serum albumin, as part of the total national program.

2. That Dr. Cohn be asked to assume general supervisory direction of the processing of albumin in commercial laboratories.

The Navy, which was primarily interested in serum albumin rather than in plasma, took the necessary steps to implement this recommendation(40).

Further recommendations were made at the Conference on Albumin on 11 February 1942 (36), as follows:

1. That the specifications for the preparation and packaging of human serum albumin prepared by Major Kendrick and Cdr. (later Capt.) Lloyd R. Newhouser, MC, USN, be accepted with certain modifications, and that future modifications be made at the discretion of these officers and Dr. Veldee.

2. That after specifications for the preparation of serum albumin were complete, Major Kendrick and Commander Newhouser meet with representatives of supply and equipment firms to determine the availability of equipment and the time required to supply the component parts for the Standard Package of Human Serum Albumin.

It was unanimously agreed, at the meeting of the Subcommittee on Blood Substitutes on 12 May 1942 (13), that the members reaffirm their earlier recommendation that serum albumin be considered as an established blood substitute of great importance and with many practical advantages. It was further recommended that clinical testing be discontinued and that the approximately 1,000 units of serum albumin in Dr. Cohn's laboratory be turned over to the Navy for clinical use.

At the third Conference on Albumin and By-Products on 26 May 1942 (40), the chairman thought it necessary, in the interests of clearness, that the membership reaffirm its recommendations and limitations by signing the following statements (which were duly agreed to and signed):6

Human serum albumin in 25% solution has been recommended to the Army and the Navy to fulfill their specific requirements for a blood substitute which can be transported and administered in small volumes with great facility and with safety. It is clearly understood, as has been etched on the containers and as is indicated in the directive to the Navy,

6This second action was taken at Dr. Cohn's insistence, because of the reluctance of some medical officers in high places to accept serum albumin.


that this concentrated solution should not be administered in severe dehydration without the simultaneous administration of fluids of such types as to maintain normal salt and water equilibria. The fluids should be given intravenously, orally if tolerated or by any other available route. The above statements apply only to the use of hypertonic albumin solution.

Human serum albumin has been established as being a blood substitute of proven value in that it causes hemodilution and raises blood pressure in a manner similar to blood plasma or serum. On the basis of clinical tests human albumin produces no more reactions than does plasma.

The stability of the human albumin solution as packaged has been established over a period of one year at room temperature in temperate climates. Such material has been carried at sea in the sick bay of a cruiser for three months at a temperature reported to average 37º C. (98° F.) without deteriorations as demonstrated by inspection and human injection. Stability has also been established by inspection at 45° C. (113° F.) for a period of 3 months and at 50° C. (122° F.) for a period of two weeks in the laboratory.


It was decided early in the albumin program that no special indoctrination of medical officers concerning serum albumin would be necessary (16). They had had no experience in its use, but numerous articles had been published on it, and the directions and warnings etched on the can were considered adequate. Informal teaching emphasized the fact that serum albumin must not be used in severe dehydration unless other fluids could be supplied to maintain the normal salt and water equilibrium. Later, a somewhat more intensive educational program was undertaken, to encourage the use of serum albumin in the Navy (7), and the Army made a filmstrip on the subject.


Although it was not legally possible to enter into contracts for the commercial production of albumin until it had become an accepted product, discussions concerning this phase of the program were recorded at various meetings of the Subcommittee on Blood Substitutes and at albumin and other conferences. Dr. Veldee took the position that the same licensed laboratories should not prepare both human plasma and albumin, but this policy was not adopted. Dr. Cohn believed it would be unfortunate to process albumin in a number of small plants and similarly unfortunate to have manufacturers build their plants on the basis of the amounts specified in the first contracts, since the Navy had promptly indicated its plans for larger amounts.

At the 18 July 1941 meeting of the Subcommittee on Blood Substitutes (6), Dr. Veldee thought that serum albumin had been proved to be of sufficient worth to be licensed as a biologic product. Dr. Cohn said that, while further improvements were to be expected, there were at this time no serious difficulties concerning the stability and sterility of serum albumin or its handling, and he foresaw no obstacles to its commercial production. Although it was against the laws of the Commonwealth of Massachusetts to prepare biologicals directly for use outside the state, the State Attorney General had ruled that


the first 4 kilos of human albumin could be thus distributed, and a bill was being prepared for liberalization of the existing law.

At this same meeting, the subcommittee recommended that all phases of commercial production be under Dr. Cohn's direct supervision. Dr. Cohn believed that the present operation at Harvard should be continued, to avert the lag which might occur when commercial houses without previous experience began to make albumin and also because the many byproducts obtained by fractionation of plasma should be developed.

All possible preliminary steps were taken as soon as the Conference on Albumin on 5 January 1942 (37) recommended serum albumin to the Armed Forces, and, at the next meeting of the Subcommittee on Blood Substitutes (12), it was reported that six commercial firms had indicated their willingness, in response to the letters of intent they had received, to process 8,500 packages of albumin each before 1 October 1942. This would mean that 180,000 additional bleedings must be absorbed by the bleeding centers. The Navy's intention to contract for 250,000 additional packages of serum albumin in the next fiscal year7 would require 750,000 additional bleedings, in addition to the 900,000 bleedings necessary for the plasma already contracted for.

Nothing came of the recommendation made at the second Conference on Albumin and By-Products on 15 April 1942 (41) that a pilot plant be established by the Army or the Navy for the processing of human and biologic products for research purposes and clinical use.

In January 1942 (37), Dr. Cohn's laboratory, Armour Laboratories, and Lederle Laboratories were considered capable of producing 1,300 units of serum albumin (25 gm. each) per week. Dr. Cohn stated that any commercial firm could participate in the program if it had large coldrooms and Sharples centrifuges and would send personnel to be trained in his laboratory for at least a month.

A number of important steps were taken after this meeting (40). Specifications for commercial production were drawn up (42). Letters of intent to purchase were sent to the various processing laboratories. Dr. Cohn received official Navy authorization to receive personnel from these firms for instruction in the preparation of serum albumin, and representatives from six producers were thus trained.

There were two chief problems connected with the program. One was the steady increase in the Armed Forces requirements for serum albumin; the requirements for fiscal year 1942-43 were set at 110,000 units for the Army and 250,000 units for the Navy. Each unit of processed albumin required 3.5 units of blood. The second problem was the delay in obtaining Sharples centrifuges. The processing of albumin, while complex, offered no particular difficulties, but the procurement of equipment to process it threatened to delay the program for many months. Dr. Cohn doubted that the amount

7Contrary to the plasma procurement policy, by which all supplies were purchased by the Army Medical Procurement Agency, Brooklyn, all serum albumin was purchased on contracts made by the Navy with processing firms. The smaller amounts of serum albumin which the Army used were purchased from the Navy.


desired could be produced within the specified time and warned that undue haste might be dangerous.

When the Conference on the Preparation of Normal Human Plasma was held at the Plasma Fractionation Laboratory, Department of Physical Chemistry, Harvard Medical School (43), the agenda included discussions and demonstrations of the size of a single batch of plasma, its electrophoretic analysis, bacteriologic filtration, preservatives, temperature stability test, coloration, hemolysis, sterility, separation of red cells, pyrogens, contamination, and plasma fractionation and processing of fractions. Clinical testing of serum albumin was no longer considered necessary. Present problems concerned quantity production and distribution. To date, there had been 41 runs at the Plasma Fractionation Laboratory, 18 at Armour Laboratories, and 6 at Lederle Laboratories.

A number of obstacles to the rapid implementation of the program were discussed, including the joint problems of securing the additional bloods (1½ million) required by the program; their cost (about $3 million per million bloods); the shortages of equipment, particularly rubber tubing; the futility of priority ratings; the possible substitution of cup centrifuges for Sharples centrifuges, which continued to be in very short supply; and the critical materials in the completed serum albumin package, which was demonstrated at the conference. It was a very different matter, Dr. Veldee remarked, to make diphtheria antitoxin from horses in the backyard and to process serum albumin from blood secured from millions of donors scattered all over the country. Dr. Cohn would not recommend that the Red Cross turn over any bloods to any processing firms until sample and practice runs could be completed with the use of hemolyzed blood or contaminated plasma, which most of these firms had on hand.

At the 23 June 1942 meeting of the Subcommittee on Blood Substitutes (14), the directive prepared by Commander Newhouser for the complete package of serum albumin was approved with the deletion of the footnote dealing with femoral injection (p. 338). At his suggestion, the label was changed to read "Standard Army and Navy Package of Normal Serum Albumin (Human, Concentrated)," the designation "serum albumin" being retained because it was customary in the chemical literature and by general usage.

A special Conference on Albumin on 9 July 1942 (44) dealt with various production matters, including procurement of materials; economy of production; length of processing; deferment of trained personnel; the risk of too great haste; and yields with various types of centrifuges, all of which were still in short supply. It was decided that the question of costs was beyond the competence of the committee, though it was brought out that they would be contingent on the rate at which the initial expenses of equipment could be written off.

At the Conference on Albumin Testing on 19 October 1942 (16), while the quantity production of concentrated human serum albumin still seemed far in the future, prospects were considerably more encouraging. The War


Production Board had expedited the procurement of necessary equipment. Practice runs with contaminated plasma had been carried out at four of the six firms to which contracts had been granted. In every instance, the chemical purity of albumin had been acceptable, less than 2 percent of globulin being present. Three firms had begun to process serum albumin from blood supplied by the Red Cross, and, after some initial difficulties from contamination, they were producing satisfactory material, which would be released after clinical testing. Personnel from Dr. Cohn's laboratory had worked in several commercial laboratories, to aid in the beginning of production. Production at the Harvard laboratory was continuing.

Up to this time (October 1942), there had been:

78 first injections of crystalline albumin with no reactions of any kind in 61.
76 second injections, with no reactions in 69.
28 third injections, with no reactions in 26.
3 fourth injections, with no reactions.

At this conference, Dr. Cohn outlined the testing of commercial serum albumin as follows:

Navy specifications called for a 25-percent solution, to save space, and for stability at 122° F. (50° C.) to avoid spoilage in tropical temperatures. As soon as a lot of albumin was filtered, six bottles were forwarded to the Harvard laboratory. To date, no reaction had occurred in patients who had received human albumin that had passed the rabbit thermal test then in use.

After Captain Stephenson and Colonel Kendrick had expressed themselves as more anxious about rubber tubing than about albumin, it was formally recommended that testing thereafter be conducted on the final package. Also, in deference to those present who thought three tests insufficient, it was recommended that clinical testing thereafter include five tests, carried out in Boston and at the Army and the Navy Medical Schools. Since a standard batch of albumin would not usually exceed 380 bottles (38,000 cc.), the revised tests would require 8 bottles, just over 2 percent, of each lot.

Progress thereafter was much faster. By the middle of December (17), four laboratories were using Red Cross blood in production, and two other firms were making practice runs. Difficulties with pyrogens had been overcome; they had been traced to filter pads and distilled water. Dr. Cohn suggested that trouble would be avoided, and valuable plasma conserved, if, from time to time, testing were carried out on distilled water, new lots of supplies, and equipment, including the intravenous equipment placed in the final Army-Navy containers.

Accelerated tests were being used, with assays carried out by electrophoretic analysis. Studies indicated that in all samples of albumin tested to date there was less globulin than the amount permitted in the specifications. Progress with all the first lots, however, was cautious, it being considered more advisable to establish correct standards than to make haste.

Requests from the Armed Forces for serum albumin continued to increase. In March 1943 (18), the Navy, which had already contracted for 360,000 units


(100,000 for the Army), asked for an additional 350,000 units, to be delivered by July 1944. Since this goal was obviously impossible with present facilities, the problem was solved by opening a new processing plant (Armour Laboratories) at Fort Worth on 1 November 1943; Fort Worth, Dallas, and Houston were the last untapped blood donor centers in the country.

At the Conference on Albumin on 22 March 1943 (18), Dr. Cohn discussed various aspects of the quantity production of serum albumin:

1. Achievement of chemical purity presented little difficulty. In all, 100 of 113 preparations examined to date had been found more than 99 percent pure, and only 1 had contained more than 1.5 percent of globulin. In view of these observations, Dr. Cohn intended to discontinue routine electrophoretic analysis of every lot.

2. Studies by Dr. George Scatchard indicated that the instability of some serum albumin preparations was caused less by the albumin than by the impurities in it. Studies on bovine albumin had contributed greatly to the stabilization of human albumin. The flocculation and haziness in certain preparations after heating remained a problem to be solved.

3. Most preparations had passed the rabbit thermal test without difficulty. After a discussion of the number of clinical tests that should be required-some observers thought as many as 12 should be used-it was agreed that the same criteria should be applied to human albumin rendered pyrogen-free by heating as would be applied to standard albumin preparations. Dr. Veldee had one reservation: Heating itself might do damage, and, for this reason, he thought that the use of heated preparations should be restricted.

4. Albumin made from contaminated plasma was apparently safe and satisfactory for use, but more caution should be used in accepting it, and more careful clinical testing carried out with it, than were employed in albumin made from uncontaminated plasma. Similar precautions should be employed in serum albumin made from plasma in broken bottles.

5. The production figures with the use of the DeLaval centrifuge were attractive, but the introduction of new methods at this stage would probably delay the program. If the program was to be expanded, and if it were certain that these centrifuges could be obtained, then it might be well to consider their use.

6. Dr. Cohn was now ready to perform only stability tests on the bulk product, and to perform all other tests in the final containers. This would save both time and material. Of the 328 clinical pyrogen tests performed between 10 June 1941 and 19 March 1943, only one preparation that had successfully passed NIH (National Institute of Health) rabbit thermal test had given any febrile reaction in man.

7. One of the commercial laboratories had been informed that a donor had developed mumps 48 hours after his donation. There was no known instance of transmission of mumps by transfusion, and it was agreed that it would not be practical to follow donors after they had given blood.

At the 9 April 1943 meeting of the Subcommittee on Blood Substitutes (45), Dr. Cohn reported that the program was proceeding satisfactorily quantitatively and so well qualitatively that a revision of the test schedule might be considered; in particular, the number of heated specimens tested could be reduced, which would save 10 days. The subcommittee authorized Dr. Cohn to establish the number and type of tests to be employed.

At the 13 May 1943 meeting (20), it was reported that 700 bottles of contaminated plasma processed into albumin by the Cutter Laboratories had been found satisfactory chemically. One of two lots tested for pyrogens had caused no reactions. The other lot had given three reactions. The present policy was to request three additional containers for examination if a reaction occurred


in any of the five routine clinical tests. If any of the three additional clinical tests was positive, the entire lot was discarded.

Delivery of serum albumin was at first disappointingly slow (table 10). Only 26,119 packages had been delivered by the end of July 1943, against an expected 150,000 (46). Then, the situation improved. Within the next 2 weeks, the number rose to 45,000 (the figures are cumulative), by the first of October (47) to more than 80,000, and by the middle of November (48) to almost 125,000. As of 2 June 1944 (23), 325,838 units of human serum albumin had been authorized for shipment to the U.S. Naval Medical Supply Depot; of this amount, 36,699 units had been produced during May. Quality continued to improve. All the albumin made at the Harvard pilot plant over the past several months had been held in the dry state, in anticipation of experimental needs.

TABLE 10.-Accepted production of units of normal human serum albumin, June 1942-November 19431

Year and month

Production laboratory





Sharp & Dohme
























































































































































1The production of the Harvard laboratory was shipped to the Bethesda Naval Hospital; that of the commercial laboratories went to the U.S. Naval Medical Supply Depot.

There had been a striking decrease in the amount of material requiring reworking because of failure to meet specifications of either sterility or stability. The yield of albumin from plasma had also increased. With modifications in


processing methods, it rose from 27.5 gm. per liter to 29.3 gm., and then to 30.2 gm. per liter. Allowing for losses, the final yield was about 25.4 gm. per liter.

After extensive testing and the ironing out of certain initial difficulties, albumin made from contaminated plasma had proved safe and effective, and sizable amounts were secured from this source. In addition, firms not making albumin were turning their contaminated stocks over to firms that did, and great savings were thus being effected.

In April 1944 (7), the Canadian National Research Council offered to the United States about 10,000 liters of contaminated serum, all of which it proposed to destroy if U.S. authorities could not reclaim it as albumin. It would be given to the Army and the Navy with no financial or other obligations. Eli Lilly and Co. were able to work the contaminated serum into acceptable, pyrogen-free serum albumin, and the Canadian offer was gladly accepted.

Some anxiety was originally felt that the development of byproducts other than albumin might slow up the albumin program. This fear was discounted. In fact, when the Navy made contracts for byproducts with firms holding contracts for serum albumin, Dr. Cohn approved the plan, on the ground that plasma fractionation was an integrated process (40).

The report of the American Red Cross Blood Donor Service to 1 September 1945 (table 11) showed that the seven commercial firms eventually involved in the albumin program had processed 2,329,175 donations into 569,014 packages, all of which had been delivered to the Navy except for 1,704 packages which one firm was holding, awaiting shipping instructions (49).


When serum albumin was in process of development, the small size of the package made it seem of great potential usefulness during landing operations and for airborne troops and such ground forces as pack-drawn mountain troops. The Army requisition for fiscal year 1942-43 was 110,000 units, but, up to 1 January 1944, because of production delays, only 36,000 units had been received. In November 1943, the requirements for calendar year 1944 were tentatively set at 150,000 units, including the 74,000 units undelivered in 1943. This requisition was later reduced to 60,000 units (50).

The reason for the reduction in the requisition for serum albumin was a revision of the premises on which the original requirements had been made (50). The small size of the package lost some of its attractiveness to the Army in the face of the necessity for making intravenous fluids available along with the albumin; unless fluids were available, albumin could not be used in dehydrated casualties. Also, experience had shown that there were almost no circumstances, including combined landing operations, in which plasma could not be supplied in adequate quantities. Albumin, of course, continued to be used according to indications in head injuries and in burns.


TABLE 11.-Summarized report of albumin production to 1 September 1945

Production laboratory


Packages delivered to U.S. Naval Medical Supply Depot





285, 809

91, 022





413, 588

101, 916





551, 143

116, 395





321, 464

79, 719




Sharp & Dohme

136, 906

26, 675





324, 766

78, 913



295, 499

74, 374



569, 014


1This firm had on hand 1,704 completed packages being held for shipping instructions.


In February 1942, when Dr. Cohn was requested to prepare specifications for the commercial production of serum albumin, he described four methods for the fractionation of plasma (40). The first technique was impractical for industry. The third, in which the supernatant of fraction IV was concentrated in a still, had to be given up because of the high incidence of pyrogens and the time required to dialyze away the accumulated salts. The fourth method, crystallization, occupied too much time. The second technique, with some modifications, was used by the processing firms. The technique is too complex to be described in detail here, but its essential steps were as follows:

1. Separation of fraction I, separation of fractions II and III, and precipitation of fraction IV.
2. Sedimentation of fraction IV.
3. Clarification of the supernatant from fraction IV, precipitation of fraction V, and reprecipitation of fraction V

More than one step could be accomplished at a time, and each could be completed within 48 hours. The complete processing of one batch would thus require a total time of 6 days. The method was the same as that employed in Dr. Cohn's laboratory; expansion of production involved no changes in the basic principles but merely changes in the mechanical equipment for handling the various steps, as well as changes in certain details, of the process. Thus, the recovery of tax-free alcohol used in the small-scale operations at the Harvard laboratory did not pay. It became an immediate problem in commercial production, and care had to be taken to provide against the distillation of volatile contaminants. There was some delay when one commercial firm was permitted to change from dialysis to capillary methods of adding alcohol. All the first material thus treated had to be reworked, but there were no difficulties in succeeding runs.



It was originally thought that albumin would require refrigeration (37). Then, it was found that it could be stored at room temperature without deterioration.8 Since, however, it was intended for use of the Armed Forces in all parts of the world, samples were tested for stability at 113° F. (45° C.) for 1 month and at 98.6° F. (37° C.) for 2 months. It was found that albumin would not remain stable at these temperatures if more than a very small amount of globulin was present. Later (34), in view of military requirements, it was decided to extend the temperature range from -58° F. to +122° F. (-50° C. to +50° C.).


Sodium Chloride

One of the first observations made about serum albumin in the Subcommittee on Blood Substitutes (34) was that it was more stable when it was made up with sodium chloride; without it, early preparations became cloudy at room temperature. The matter was to come up in other meetings of the subcommittee and at various conferences on albumin.

An ad hoc committee (Dr. Veldee, Dr. Earl S. Taylor, and Lt. William G. Workman, MC) was appointed to study the problem in May 1942 (13), after Dr. Cohn stated that the addition of sodium chloride to the blood intended for plasma fractionation greatly complicated the processing of serum albumin. The higher the salt content, the larger was the amount of globulin passing into fraction V, and the larger the amount in fraction VI. The ad hoc committee found the plasma yield to vary by less than 1 percent with and without salt. These observations paralleled those made by Eli Lilly and Co. It was also found that the total osmotic pressure of 25-percent albumin could be considerably increased by adding more sodium chloride and considerably decreased by reducing the amount used if either change were desired(16).

In January 1943, at a meeting of the Subcommittee on Albumin and By-Products (51), it was pointed out that the clinician's preference for isotonic solutions of serum albumin might be on unsound grounds from the standpoint of physical chemistry; a hypertonic solution would increase the stability of the product. It was proposed that the sodium chloride content of serum albumin be increased to 2 percent. The amount of solution injected was so small that there could be no valid objection to this increase. Potassium salts, however, should not be added, because some of the clinical conditions encountered would be complicated by hyperpotassemia.

8At this time (1962), albumin is stored at 4° to 6° C., and the dating period has been correspondingly increased.


In March 1943, the salt content of serum albumin was altered from 0.15 to 0.3 molar. The change at least doubled the stability of the product at 135° F. (57° C.) and also increased its stability at 122° F. (50° C.), though less strikingly (48). Samples from each of the processing firms kept for 100 days at 122° F. (50° C.) looked clear enough at the end of the period to be used clinically (table 12). Dr. Cohn would have expected a twofold improvement in the stability of the product with the doubling of the salt component, not the sixfold increase that had occurred and that might reflect increased skill on the part of the producers.

When 0.3 molar sodium chloride was first used, authorization was given for making as much as 0.09 of this component sodium acetate. Later (52), Dr. J. Murray Luck and his associates at Stanford University demonstrated that a further increase in the length of the paraffin chain of the anion would increase for this purpose the stability of the serum albumin. Sodium butyrate was tested and discarded because of the undesirable odor. Albumin was then prepared with 0.05 molar sodium phenylacetate plus 0.25 molar sodium chloride. Still later, sodium mandelate was substituted for phenylacetate because of its therapeutic properties. Tests showed that whichever of these agents was used would make possible the heating of albumin for hours at temperatures close to 168° F. (70° C.), a step which would also reduce the dangers of bacterial or virus contamination (tables 12 and 13).

TABLE 12.-Effect of storage at various temperatures on stability at 57° C. of crystalline human albumin (lot HA-64)

Conditions of experiment

Hours required for increase
of 50 Mueller units

Hours required for 5%
increase in viscosity



















Storage days

Storage temperature











Degrees C.








































































TABLE 13.-Effect of storage at various temperatures on viscosity of crystalline human albumin (lot HA-64)

Conditions of experiment

Ratio of viscosity after storage to initial viscosity











Storage days

Storage temperature





Degrees C.












































In a supplementary report, Dr. Luck and his associates (53) recommended that a large-scale clinical experiment be conducted with 25-percent human serum albumin solution, of pH 6.6 to 7.0, containing 0.04 molar sodium mandelate and 0.26 molar sodium chloride. Such solutions would probably be of great stability and, if clinical trials revealed no adverse effects, should be employed in place of the present serum albumin solution. Dr. Hans Clarke had suggested acetyl phenylalanine, and it was thought that it might be possible to replace mercurials with it in both albumin and dried plasma. Studies based on these proposals were limited to trial runs.

Meantime, thermal stability was generally improving, and by the spring of 1944 (52), it was unusual to find more than an occasional sample with stability of less than 50 hours at 135° F. (57° C.). Mueller nephelometers had been distributed to the processing houses, and comparative studies of their results and those of the Harvard laboratory showed satisfactory agreement. Electrophoretic and ultracentrifugal controls had long since been discontinued.

In October 1944, Captain Newhouser directed all laboratories fractionating plasma under Navy contracts to prepare serum albumin of low salt content "if such a change is commercially feasible, will supply a product with satisfactory stability and does not increase the contract cost of the serum albumin."

Isoelectric albumin could be prepared practically free of sodium ions, but it was stable only in the dry state and could not be dispensed in solution. If there were sufficient advantage in preparing completely salt-free albumin, it could be dispensed in the dry state, with glucose in the diluent bottle, but this


would require new contract specifications and a new package. No clinical evidence existed, however, of the value of reducing the sodium ion concentration below a certain level.

A preliminary report of this investigation was made on 16 January 1945 (54), and a summary report was made to the Subcommittee on Blood Substitutes on 16 March 1945, by Dr. Cohn (55), as follows:

1. Accumulated evidence to date indicates no significant behavioral differences between the new salt-poor albumin and the standard salt-containing preparation in shocked or in normal individuals. In shock, injection of 25-percent albumin, whether salt-poor or salt-containing, always increases the blood volume. In dehydration, the injection of additional fluids is still recommended.

2. Preliminary evidence indicates that salt-poor albumin is an effective diuretic agent in the nephrotic syndrome and in hepatic cirrhosis associated with severe edema.

3. In a small group of surgical patients with hypoproteinemia and edema, concentrated albumin has been found to mobilize water from the interstitial spaces into the plasma. Used in a few patients after trauma, it produced a significant rise in plasma protein concentration.

4. Salt-poor albumin, stabilized with 0.04 molar acetyl tryptophan or 0.04 molar mandelic acid, retains its stability better than standard serum albumin preparations when heated at 140° F. (60° C.) for 10 hours. This procedure, or some comparable heating procedure, would destroy most bacteria in vegetative forms as well as such viruses as that of infectious hepatitis. It has been recommended that studies be undertaken as soon as possible of the time required to destroy the agent of hepatitis at a given temperature.9

5. The original opinion of serum albumin as to its convenience, rapidity and ease of administration, and effectiveness in increasing plasma volume in injury and shock has been fully confirmed by this investigation.

As a result of the evidence secured in this investigation, it was recommended that salt-poor albumin be substituted for the current preparation, which contained 0.3 molar sodium chloride. It should be prepared without a mercurial preservative (p. 354); should be stabilized with 0.04 molar acetyl tryptophan, a derivative of a natural amino acid; and should be heated for such periods and at such temperatures as might be necessary for the destruction of viruses.

At this meeting it was also recommended that products of plasma fractionation, including human serum albumin, be included in the "United States Pharmacopeia." This recommendation was not made effective until November 1950, with the 14th edition. Plasma had first appeared in the 1942 edition.

Antibacterial Agents

At the January 1942 Conference on Albumin (37), Dr. Cohn reported that the experience gained at one of the processing companies had indicated that it was highly dangerous to add Merthiolate to albumin. Experimental studies were to the same effect. The production of albumin was delayed at least 6 weeks by the necessary investigations. A proposal that preservatives be eliminated entirely was made at the 19 July 1943 conference (56), but if this was

9The menace of hepatitis was just beginning to be appreciated (p. 647).


done, it was pointed out in the discussion, at least a third of the total product would have to be tested. This plan would be workable as far as individual bulk containers were concerned but totally impractical in the final containers.

The problem, as Dr. Cohn pointed out at this conference, fell into two parts, (1) filtration to achieve a sterile product and (2) the use of preservatives. With a 30-percent solution, the viscosity was too great for convenience. Later (16), Dr. Cohn reported an observation not previously made, that 25-percent normal human serum albumin is essentially isoviscous with respect to blood.

Serial studies by Col. Elliott S. Robinson, MC, showed that the more rapidly the albumin was filtered, the better the results from the standpoint of sterility. Using Seitz filters, he found that contamination was not a particularly troublesome problem. He attributed the good results partly to good luck and partly to attempts to expedite the procedure. Filtration became easier as the globulin content of preparations was reduced.

Studies reported by Dr. Janeway at the 19 July 1943 meeting (56) had been undertaken on the assumption that alcohol was a good bacteriostatic agent. In four complete runs, and an almost completed fifth run, he had taken bacterial counts at every fractionation and found contamination low; the average, figured in terms of the original plasma to take care of dilution, was 30 colonies per cc. Although the counts were very low, sterility was not achieved at any point until fraction V, which carried down bacteria, was taken off. The supernatant was free from bacteria, but the precipitate was not; it usually contained a smaller number of bacteria than might have been anticipated, but the counts might run up to about 100 colonies per cc. of redissolved fraction V. Samples were taken in quadruplicate. The results were the same whether they were frozen or left at ice box temperatures, but there was occasionally a considerable difference between cultures made at room temperature and at 98.6° F. (37° C.).

It was concluded that alcohol did not act as a sterilizing agent, at least at the temperatures at which these studies were conducted. It was also concluded that bacteria were collected at most of the steps in the process, from the operators, filters, and other sources, but that the number was probably not great enough to have any influence on pyrogens, though if the organisms were of the gram-negative variety, a few would be sufficient to cause trouble.

Dr. Veldee agreed that, assuming dried serum albumin to be relatively sterile, the important points would be the speed of getting the material into solution, the temperature of the solution, and the speed with which it went through the filter. Colonel Robinson's experience had been limited to relatively small lots. The NIH experience had included 75-liter lots. Colonel Robinson had used a very simple type of Seitz filter, which could be cleaned and inspected for leaks. Dr. Veldee doubted that such filters would be used in the larger commercial houses and thought that in certain circumstances the technique of filtration might well be a source of contamination. If there was a small defect in the filter, there was more chance of its showing up with 75- than with 25-liter lots.


As to preservatives, Merthiolate in 1:15,000 concentration had always been acceptable to NIH, but if 1:10,000 had been proposed when the minimum requirements for human plasma were being drawn up, it would have been equally acceptable. The problem was this: In all other biologics except human plasma and serum, the dosage was small, usually 1 or 2 cc., or under 10 cc., and the effective bactericidal dose per patient was equally small. Therefore, any one of a number of preservatives could be used to kill any living organism. With albumin, plasma, or serum, however, such large doses had to be given that preservatives could not be used in the same concentration as in smaller dosages because of the risk of toxicity.


At the Conference of the Albumin and By-Products Group on 14 December 1943 (57), Dr. Cohn pointed out that the development of serum albumin had undergone two phases. In the first, all the material produced was devoted to an appraisal of its efficiency in shock. In the second, all the material was delivered to the Armed Forces for use overseas. At this time, with full commercial production underway, the output of the pilot plant at the Harvard Medical School constituted less than 1 percent of the total production, and he believed that a third phase might be considered, in which the output at this plant could profitably be devoted to the study of diseases and conditions other than shock. He agreed with those who pointed out that, while the amount of serum albumin thus used would be small in comparison with the total amount, great quantities would be required in the treatment of nephrotic states, cirrhosis, and nutritional edema, and the needs of the Armed Forces must not be jeopardized in any way by the proposed research program.

In December 1944, the Committee on Medical Research, NRC, was requested by the Navy to provide additional clinical data on serum albumin, with special reference to the increase in plasma volume accomplished by its use with and without the additional administration of crystalloid solution. An additional purpose of the investigation was the testing of the new salt-poor albumin preparation just developed. The conditions studied included trauma of various kinds, with and without shock and with and without active fluid loss at the time of the investigation; postoperative states, with and without shock; and medical conditions. Normal subjects were also studied.

Considerable scattering was found in the increase in plasma volume per gram of albumin injected, which was to be expected, even in normal subjects, as the result of differences in tissue hydration, circulatory state, renal activity, and other variants. Injured subjects, especially those with hemorrhage, burns, or peritonitis, who were losing fluid actively, sometimes held little or none of the injected albumin and fluid and showed only insignificant increases in plasma volume.


The following conclusions were considered warranted:

1. As an overall average, the injection of concentrated human serum albumin was associated with a rapid, and sometimes immediate, increase in plasma volume of 12-14 cc. per gram of injected albumin. The same average held per gram of albumin retained.

2. Administration of a fixed amount of saline solution intravenously with the serum albumin (800 cc. with 200 cc. of 25-percent albumin) resulted in an appreciable increase in plasma volume as compared with the increase that followed the administration of albumin alone. Experimental studies on dogs, in which dehydration was produced by withholding fluids or by inducing diuresis with glucose, had shown that the administration of albumin alone did not restore the circulating blood volume and blood pressure to normal and that the survival rate was low.

3. Patients in severe shock, presumably with continuing losses of blood or plasma, showed much smaller increases in plasma volume per gram of albumin given (an average of 8 cc. in shock against 14 cc. without shock). Thus, in severe shock, the administration of 25 gm. of serum albumin without additional fluid would correspond to a plasma infusion of 200 cc. Maj. (later Lt. Col.) Henry K. Beecher, MC, reached approximately the same conclusion in his study of battle casualties in Italy, although the only criterion of improvement available to him was arterial blood pressure (p. 40).

4. In several instances, the administration of additional intravenous salt solution after the administration of albumin produced further increases in plasma volume.

5. No differences were noted in the results of salt-poor and salt-containing solutions.


The production of serum albumin consumed an increasing number of blood donations until February 1944, when about 30,000 bloods a week were being supplied for this program. Then, with the needs of the Navy and the far smaller needs of the Army well provided for and supply simply a maintenance matter, production was gradually reduced until 15 October 1944, when it was cut sharply. The Armour Laboratories plant in Fort Worth was closed and the blood donor centers which had supplied it, in Fort Worth, Dallas, and New Orleans, were also closed. Four other laboratories ceased receiving blood for serum albumin at this time, but other processing laboratories continued to operate, on a curtailed scale, until the end of the war.


1. Wangensteen, O. H., Hall, H., Kremen, A., and Stevens, B.: Intravenous Administration of Bovine and Human Plasma to Man: Proof of Utilization. Proc. Soc. Exper. Biol. & Med. 43: 616-621, April 1940.

2. Minutes, meeting of Committee on Transfusions, Division of Medical Sciences, NRC, 31 May 1940.

3. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 30 Nov. 1940.

4. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 19 Apr. 1941.

5. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 23 May 1941.

6. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 18 July 1941.

7. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 21 Apr. 1944.


8. Minutes, Conference on Bovine Albumin, Division of Medical Sciences, NRC, 16 July 1942.

9. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 20 Oct. 1942.

10. Kremen, A. J., Hall, H., Koschnitzke, H. K., Stevens, B., and Wangensteen, O. H.: Studies on the Intravenous Administration of Whole Bovine Plasma and Serum to Man. Surgery 11: 333-355, March 1942.

11. Report, Subcommittee for the Standardization of Dispensing Equipment, Committee on Blood Substitutes, Division of Medical Sciences, NRC, 8 May 1941.

12. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 10 Mar. 1942.

13. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 12 May 1942.

14. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 23 June 1942.

15. Minutes, meeting of Committee on Medical Research, Division of Medical Sciences, NRC, Office of Scientific Research and Development, 7 July 1942.

16. Minutes, Conference on Albumin Testing, Division of Medical Sciences, NRC, 19 Oct. 1942.

17. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 15 Dec. 1942.

18. Minutes, Conference on Albumin, Division of Medical Sciences, NRC, 22 Mar. 1943.

19. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 23 Mar. 1943.

20. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 13 May 1943.

21. Edwards, F. R.: Despeciated Bovine Serum (D.B.S.): A Substitute for Human Plasma. Brit. M. J. 1: 73-76, 15 Jan. 1944.

22. Wangensteen, O. H., Kremen, A. J., and State, D.: The Experience of the Surgical Clinic of the University of Minnesota Medical School in the Use of Bovine Albumin as a Blood Substitute. Committee on Medical Research, Office of Scientific Research and Development, 10 Sept. 1943.

23. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 2 June 1944.

24. Heyl, J. T., Gibson, J. G., 2d, and Janeway, C. A.: Studies on the Plasma Proteins. V. The Effect of Concentrated Solutions of Human and Bovine Serum Albumin on Blood Volume After Acute Blood Loss in Man. J. Clin. Invest. 22: 763-773, November 1943.

25. Mann, F. C.: Further Experimental Study of Surgical Shock. J.A.M.A. 71: 1184-1188, 12 Oct. 1918.

26. Strumia, M. M., Wagner, J. A., and Monaghan, J. F.: The Intravenous Use of Serum and Plasma, Fresh and Preserved. Ann. Surg. 111: 623-629, April 1940.

27. Elliott, J.: Preliminary Report of a New Method of Blood Transfusion. South. Med. & Surg. 98: 643-645, December 1936.

28. Stokes, J., Jr., Mudd, S., Roddy, R. L., Eagle, H., Flosdorf, E. W., and Lucchesi, P.: The Use of Lyophile Human Serums for Prevention and Treatment of Infectious Diseases. Am. J. Dis. Child. 48: 1428-1429, December 1934.

29. Flosdorf, E. W., and Mudd, S.: Procedure and Apparatus for Preservation in "Lyophile" Form of Serum and Other Biological Substances. J. Immunol. 29: 389-425, November 1935.

30. Flosdorf, E. W., and Mudd, S.: An Improved Procedure and Apparatus for Preservation of Sera, Microörganisms and Other Substances-the Cryochem-Process.  J. Immunol. 34: 469-490, June 1938.

31. Fantus, B.: The Therapy of the Cook County Hospital. Blood Preservation. J.A.M.A. 109: 128-131, 10 July 1937.


32. Bond, D. B., and Wright, D. G.: Treatment of Hemorrhage and Traumatic Shock by the Intravenous Use of Lyophile Serum. Ann. Surg. 107: 500-510, April 1938.

33. Mahoney, E. B.: A Study of Experimental and Clinical Shock With Special Reference to Its Treatment by the Intravenous Injection of Preserved Plasma. Ann. Surg. 108: 178-193, August 1938.

34. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 19 Sept. 1941.

35. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 3 Nov. 1941.

36. Minutes, Conference on Albumin, Division of Medical Sciences, NRC, 11 Feb. 1942.

37. Minutes, Conference on Albumin, Division of Medical Sciences, NRC, 5 Jan. 1942.

38. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 24 Feb. 1943.

39. Minutes, Conference on Albumin, Division of Medical Sciences, NRC, 11 May 1942.

40. Minutes, Conference on Albumin and By-Products, Division of Medical Sciences, NRC, 26 May 1942.

41. Minutes, Conference on Albumin and By-Products, Division of Medical Sciences, NRC, 15 Apr. 1942.

42. Specifications for a Complete Package of Serum Albumin (Human) Concentrated for the Army and Navy, n. d.

43. Minutes, Conference on the Preparation of Normal Human Serum Albumin, Division of Medical Sciences, NRC, 5-6 June 1942.

44. Minutes, Special Conference on Albumin, Division of Medical Sciences, NRC, 9 July 1942.

45. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 9 Apr. 1943.

46. Minutes, Conference of Albumin and By-Products Group, Division of Medical Sciences, NRC, 28 July 1943.

47. Minutes, meeting of Subcommittee on Blood Substitutes with Subcommittee on Neurosurgery; Conference of Albumin and By-Products Group, Division of Medical Sciences, NRC, 5 Oct. 1943.

48. Minutes, Conference of Albumin and By-Products Group, Division of Medical Sciences, NRC, 17 Nov. 1943.

49. Final report, Albumin Production, American National Red Cross Blood Donor Service to 9/1/45.

50. Memorandum, Lt. Col. Douglas B. Kendrick, MC, for Brig. Gen. Fred W. Rankin, 2 Feb. 1944, subject: Requirement for Human Serum Albumin.

51. Minutes, Conference on Production of Normal Human Serum Albumin and Its By-Products, Division of Medical Sciences, NRC, 22 Jan. 1943.

52. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 2 Mar. 1944.

53. Luck, J. M., Ballou, G. A., Bassett, D., Boyer, P. D., Lum, F. G., and Rice, R.: The Stabilization of Standard Human Serum Albumin With Sodium Mandelate and Sodium Phenylactetate. Blood Substitutes Report No. 15, Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 15 Mar. 1944.

54. Salt-Poor Albumin. Interim Report, under Contract OEMcmr-139, Department of Physical Chemistry, Harvard Medical School, 16 Jan. 1945.

55. Minutes, meeting of Subcommittee on Blood Substitutes, Division of Medical Sciences, NRC, 16 Mar. 1945.

56. Minutes, Conference on the Production of Normal Human Serum Albumin and Its By-Products, Division of Medical Sciences, NRC, 19 July 1943.

57. Minutes, Conference of the Albumin and By-Products Group, Division of Medical Sciences, NRC, 14 Dec. 1943.