Studies of the Recovery and the Cost of Low-Glycerol Cryopreserved Human Red Blood Cells H. S. BOWMAN, F. A. OSKI,J. REIHART, M. A. SIMMONDS, A N D R. K. CUNNINGHAM From rhe Hemarology Cenrer. Harrisburg Hospiral. Harrisburg. Pennsylvania, and the Pediarric Reseorch Lobororory. Upsrare Medical Cenrer. Svrocute. New York

preservation, recovery, and transfusion into human recipients. This method results in excellent recovery of red blood cells, and provides rapid and efficient deglycerolization at a somewhat lower cost.

Red blood cells were equilibrated with 28 per cent ( v / v ) glycerol and 3 per cent mannitol in 0.65 g/IOO ml sodium chloride. 'The units were frozen by immersion into liquid nitrogen and stored at - 160 C. After thawing, they were reconstituted and washed wing the IBM 2991 Blood Cell Processor. Freezethaw rate C U N ~ ~ ,the effect of thawing techniques, the effect of varying postthaw washing and processing techniques, estimates of red blood cell loses because of hemolysis, and in virro recovery were determined. In vivo recovery was determined by 5'Cr techniques 24 hours after infusion and Ashby survivals and subsequent life span were measured. Metabolic, scanning electronmicroscopy, cost estimates and quality control studi&. were done on the reconstituted red blood cells. Recipients were evaluated before and after transflsion for metabolic .erythrocyte characteristics and for evidence of hemolysis. The modified method requires less wash solution and less technician time than does the standard low-glycerol method. Two units for the same recipient could be passed through the IBM software with no alteration of cell survival or loss. Revision of the IBM 2991 processing procedure provided excellent recovery of viable previolsly frozen red blood cells at probably a lower cost.

Materials and Methods Donors and Recipients Blood was collected from 42 volunteer donors into citrate-phosphate-dextrose (CPD). After collection of 450 ml blood in 63 ml of CPD, the units were centrifuged and the 'plasma was removed. The volume of packed cells was determined by dividing the net weight of the cells by their specific gravity. From 210 to 245 ml (mean235) of packed cells were obtained from each unit. Forty-two recipients were used, all with obtained informed consent. A few recipients had posthemorrhagic anemia in a steady state, but most recipients were healthy. The units used for these studies were selected from a total of 1,213 processed by the low-glycerol method and transfused during the past three years. Thirtyeight persons received autologous or ABO and Rh,(D) compatible heterologous red blood cells. Four Group A individuals were transfused with Group 0 red blood cells and were used for studies of red blood cell survival by the Ashby method. These and healthy volunteers used for "Cr studies were phlebotomized 500 ml prior to transfusion. The amount of cells transfused was less than 300 g. Total postthaw processing time was about 20 minutes (one unit) and 45 minutes (two units).

THECRYOPROTECTIVE EFFECT of glycerol was demonstrated by Smith.24 Glycerol protects red blood from damage caused by freezing and thawing and permits low temperature frozen storage. Many suitable methods have been devised for red blood cell cryoprotection, reconstitution, and successful tranSfuSion.8.12.13.IS. 19.20.22.25.27 A primary disadvantage of frozen red blood cells is the high cost of processing and storage. In this paper, we wish to describe another method for low glycerol red blood cell cryo-

Processing and Storage Equal volumes of 28 per cent (vfv) glycerol with 3 per cent mannitol in 0.65 g/100ml sodium chloride were allowed to equilibrate with the red blood cells upon a shaker at slow speed (30 cycles/minute) for ten minutes. Following an additional incubation for 20 minutes at room temperature, the glycerolized cells were transferred to a polyolein freezing bag (UCAR #7430-

Received for publication February 14, 1975; accepted J u n e l l , 1975. These studies were supported by both Grant X-5, Harrisburg Hospital Educational and Research Foundation and by U.S. Public Health Service Grant HD-

01919.

1 I3 Transfusion Mar.-Apr. 1916

Volume 16 Number2

114

BOWMAN, ET AL.

3)*. All air was expressed from the bag in sterile fashion and the port was sealed electronically with a Linde Port Sealer.* The top of the pack with outlet ports was carefully folded and the unit placed inside an anodized aluminum container* which is designed to maintain the thickness of the blood film a t about 7 mm during freezing. The red blood cells usually were frozen within 24 hours of collection by complete immersion of the container in liquid nitrogen. The aluminum container was locked within a freezing press** to maintain it upright, rigid and immobile. The freezing was considered to be complete when the nitrogen ceased to boil. Temperature of some units were derived from thermocouple recordings. The containers were stored at -160 C in twothirds liquid and one-third vapor phase of liquid nitrogen, within round liquid nitrogen refrigerators (LR-310*** and MVE-4500t). Thawing Techniques The units were thawed rapidly in a 45 C water bath. In this modified method, the aluminum container was removed from liquid nitrogen and plunged into the water bath and agitated by “edging” once per second in the vertical position, for one minute. The unit was turned to a horizontal position, one plate of the aluminum container was opened, and the unit agitated in a flat position. The bottom side of the container was used as a tray. The cells thus could be examined visually and manually to assure full and complete thawing. Then the cells were removed from the bath and the temperature of an aliquot was recorded. Freeze- Thaw Temperatures Six units of packed glycerolated red blood cells processed were monitored during freeze-thaw processing by a calibrated copper constantan thermocouple placed into the center of the blood film within a polyolein pack. Temperature during freezing and thawing was recorded every five seconds.$ These units were not used for transfusion. Post t ha w Processing With one minute’s constant mixing, 500 ml buffered 3.2 per cent sodium chloride$ was added ‘Union Carbide, Chicago, Illinois. **Cryogenic Supplies, Buckeystown, Maryland. ***Union Carbide. Chicago, Illinois. +Minnesota Valley Engineering, Prague, Minnesota. $West Guardsman Recorder, Phoenixville, Pennsylvania. SCytosol Laboratories, Boston, Massachusetts

Transfusion Mar.-Apr. 1976

to the thawed red blood cells before the unit was attached to the disposable plastic harness and inserted into the circular centrifuge bag within the IBM 299 I blood processor. After two centrifugations of two and one-half minutes each at 2990X g , all the supernatant fluid was automatically pumped from the centrifuge bag into a waste pack. The cells were then washed twice with 300 ml buffered 0.9 per cent saline with 0.2 per cent dextrose and centrifuged. After a final wash with 400 ml similar solution all supernatant fluid was pumped into the waste pack. The total volume of the supernatant fluid and hemoglobin in the pack was measured. The washed reconstituted red blood cells were removed from the centrifuge and drained into a 600 ml transfer pack Qby gentle kneading. The harness and all sterile tubing were retained in the IBM processor and the almost empty centrifuge pack was replaced in the centrifuge cage. A second unit of red blood cells was then deglycerolized and washed in the same manner. These cells were added to the transfer pack with 50 ml of fresh C P D plasma. The processed units had a shelf life of 24 hours. IBM 2991 Blood Cell Processor The IBM 2991 Blood Cell Processor is a compact unit containing a semi-automatic centrifuge and a hydraulic pump to push off the supernatant fluid. A disposable software harness connects the central bag in the centrifuge with containers of thawed red blood cells, wash solutions, and a waste bag. The equipment can be programmed in advance to carry out the wash cycle automatically in proper sequence for batch washing. With this equipment, a unit of low-glycerol thawed red blood cells may be deglycerolized with I .5 liters o f wash solution in about 20 minutes. 5 1 C rLabelling and Blood Volume One hundred microcuries of 51Crwere added aseptically to 25 packed cell units, prior to glycerol addition and f r e e ~ i n g .The ~ . ~unit ~ was gently mixed at room temperature for 45 minutes and then 1 ml of ascorbic acid (50 mgs) was added to reduce the hexavalent radiochromate to its inactive form.8 The unit was then processed as previously described. Aliquots for 51Cractivity were taken from the unit immediately prior to freezing, postthaw, and of the final reconstituted preparation for transfusion. Blood samples were obtained from recipients 24 hours after transfusion, and then usually daily for 14 to 28 days. All samples were counted simultaneously for gamma a ~ t i v i t y . ~ .Units “ were usually stored 14

n Travenol Laboratories, Morton Grove, Illinois.

Volume 16

Number2

LOW-GLYCEROL CRYOPRESERVED RED BLOOD CELLS

days before transfusion so the recipients received 50 pc or less of "Cr as a result of isotopic decay, and isotope loss during thawing and deglycerolization. Packed cell volume was determined on each equilibration sample and 51Cr activities were measured as cpm/ml rbc.I4. When two red blood cell units were transfused, only one was tagged with 51Cr. Ten individuals received 51Cr autotransfusions. In these, 5LCrradioactivity was determined 24 and 48 hours after infusion. In 31 recipients receiving reconstituted red blood cells prepared by this method, 1251-h~manserum albumin was used to determine plasma volume. The red blood cell volume was estimated from the lzSIplasma volume and the total body hematocrit as d e s ~ r i b e d . ~Prior . ~ ~ to infusion of labeled material, 500 ml of blood was withdrawn from the recipient. As the red blood cell volumes so derived for single unit transfusions agreed within 45 ml of 1.1 mg/kg," this estimates based upon 28 ml measurement was assumed accurate. In double unit transfusions, the derived red blood cell mass was corrected by the increase ml of red blood cells administered with the second unit.

*

In Vivo Survival The immediate posttransfusion 100 per cent retention value (U,Ro) was calculated as described p r e v i o ~ s l y . The ~ 51Cr 24 hour unit activity (cpm/ml rbc) divided by U, R, was equal to the percentage 51Crsurvival at 24 hours. The Cr regression curve was followed for 14 to 28 days, and its slope derived by a method of least squares.14 Four Group A recipients of reconstituted Group 0 red blood cells were studied by an automated Ashby differential agglutination procedure to derive both 24-hour survival and subsequent mean cell life.

In Vitro Recovery Total cellular hemoglobin postthaw was determined by cyanmethemoglobin and divided into the total supernatant hemoglobin to yield the percentage of freeze-thaw hemolytic loss, multiplied by 100. Supernate hemoglobin was measured by a benzidine method.2 Volume estimates of all preparations were made from net weight divided by specific gravity. The overall recovery of red blood cell mass in vitro was calculated by dividing the total residual cellular hemoglobin (g) by the original total cellular hemoglobin (g) postthaw, multiplied by 100. Leukocytes in final red blood cell preparations were counted in duplicate in a Coulter Model S electronic counter. Index of Therapeutic Viability The index of therapeutic viability was the product of percent total hemoglobin yield in the

115

final reconstituted preparation multiplied by the 24-hour 51Crper cent s ~ r v i v a l . ~ ~ ~ ~ ~ Melabolic and Hemolytic Studies Prior to glycerolization, aliquots from random donor units were removed and potassium, rate of glycolysis,16.1iand measurements of 2,3-DPG and ATP were obtained. These studies and pH were repeated on the final red blood cell unit. Baseline pretransfusion values for serum potassium, 2,3-DPG and ATP were obtained from recipients. These were repeated at one and 24 hours posttransfusion.11.16.6Five random units were studied for glycolytic intermediates by methods which have been pub1ished.l' Prior to the addition of glycerol, hematocrit, hemoglobin, plasma hemoglobin, serum haptoglobin, lactic dehydrogenase, and bilirubin were determined by standard method^.^ These were repeated upon the final red blood cell unit to be transfused, and on samples obtained from all recipients before transfusion. They were repeated at 30 minutes, one hour, 24 hours, and for up to seven days after infusion of the reconstituted frozen red blood cells. Estimates of Cost and Quality Control Cost accounting statistics' were used to estimate the cost for frozen red cell processing by this method. The approximate costs for expendable items such as software and,solutions, and technologist's labor during freezing and thawing are shown (Table 2). The labor cost factor is based upon actual time spent by the technologist responsible for the IBM 299 1 processor at a rate of $4.00 per hour. The total estimate is based upon using one IBM washing pack harness for sequestial processing of two units of red blood cells and includes time for examination of the software for any seam leaks during single o r double usage. Equipment and storage costs were not included. These costs depend on the procedures selected and the quantity of units frozen, stored, and processed within different laboratories.21 51Crdata was used to measure cellular carry over between the first and second unit. Only unit one was labeled with radiochromium. Final reconstituted red blood cell preparations were required to have a supernate hemoglobin of 200 mg or less, and were cultured for aerobic and anerobic microorganisms, for at least one week. Posttransfusion recipient sampling as outlined served as added in vivo quality control. Several red blood cell preparations were appropriately fixed in alcohols' and examined by scanning electron-micrography .

116

Transfusion Mar.-Apr. 1976

BOWMAN, ET AL. 40

+ 20

0.

-20 40

60

FIG. I . Thermocouple readings of the freezing rate and the thawing rate and unit temperatures, using polyolein packs and aluminum holders.

80

1.00

I20

140

160

I80

re c o n d r

1

f

1

10

20

30

1

40

( 1.7O

1

1

I

1

SO

60

70

80

C / sec 1

( l.8'C/

sec.)

l

-

90

1

1

1

100

110

120 130 140

1

1

FREEZING THAWING

Results Freeze-Thaw Rates Initial prefreezing rates fell gradually, but below -3 C were rapid and linear. Freezing was complete at 120 to 125 seconds, and showed no eutectic point. Reciprocal thawing curves measured a rapid initial component with uniform and complete thawing at 120 seconds (Fig. 1).

In Vivo Survival The mean 24-hour survival and recovery was 93.5 per cent for double sequential units, and 92.7

per cent for single units. These results are not statistically different. Survival curves exhibited a n initial fall from the 15-minute immediate posttransfusion 51Cr activity to the 24-hour value. After this, the regression curves had a normal slope as followed to 28 days. The mean in vivo half life of 5'Cr tagged red blood cells was 27.9 1.8 days. Survival curves from the four individuals that received these processed red blood cell units are shown in Figure 2. 10 subjects where the 24-hour red blood cell recovery was derived from red cell volume estimates and surface areas, the mean 51Crvalue was 93.1 per cent. The mean

*

Volume 16 Number 2

LOW-GLYCEROL CRYOPRESERVED RED BLOOD CELLS

117

0

LOG -1.

51 C t

I

7 1 2 . 2 9 . 9 O&YS

l t 2 . 2 6 9 DAYS

0

21

FIG.2. Representative Cr survival regression curves for four transfused recipients of low-glycerol reconstituted red blood cells by this modified method.

24-hour survival values determined by the Ashby method was 92.1 per cent. Mean cell life spans ranged from 93 to 124 days. In Vitro Red Blood Cell Recovery The mean total supernatant hemoglobin in 65 units processed by this method was 3280 -f 1.29 mg. The freeze thaw total recovery thus was 95.3 + 1.6 per cent (Fig. 3). The red blood cell recovery as measured by hemoglobin mass was derived in 30 double units and 12 single units. Total recovery, including in-line cellular processing losses, was 90.1 and 90.6 per cent for double and single units respectively (Fig. 4). Leukocyte Levels and Index of Therapeutic Viability Leukocytes in the final preparation ranged from 50 to 1100 leukocytes/mm3(mean 440). The majority of these were lymphocytes. For double sequentially processed units and for single units, the index of therapeutic viability was 84.2 and 83.9 per cent respectively (Normal 79.5% or greater).25

Metabolic Studies Red blood cell potassium was not remarkably reduced in the reconstituted type units. Cellular potassium ranged from 97.1 to 101.5 mEq/ml rbc. The mean pH of the reconstituted units was 6.9. Concentrations of ATP were normal when determined immediately after thawing and processing of red blood cells. Some units (Table 1) did have a slightly reduced glucose consumption as lactate production was minimally diminished. Postthaw units and those in the final preparation had somewhat decreased 2,3-DPG levels. The initial .value was 4,305 units and 3,555 units were in the posttransfusion preparation. Glucose-6-phosphate tended to be slightly low, but there was no accumulation of total triose phosphate (TTP) or other intermediates. Measurements in Recipients No increases in plasma hemoglobin or lactic dehydrogenase were seen in any volunteers during the initial 60 minutes posttransfusion. No abnormalities of renal function, bilirubin, haptoglobin,

118

Transfusion Mar.-Apr. 1976

BOWMAN, ET AL.

TOTAL SUPERNATE H EYOGLOBI N

tmq)

N*65

5000

t

-

90 80

-

70

3000

-

60 50

2000

-

4000

40

FIG. 3. Plots of total supernatant hemoglobin losses from freezing and thawing, and of the total in vifro per cent recovery after processing in the IBM 2991 apparatus.

30 I000

20

10

serum potassium, or urine hemosiderin accumulation were detected, and no hemoglobinemia occurred. At 24 hours, red blood cell ATP and 2,3-DPG in recipients were normal. Estimates of Cost and Quality Control Cost estimates are found in Table 2. The expense using the IBM 2991 processor is based upon double sequestial blood units with a single IBM software harness, preparing two sedimented red blood cell units for the same recipient. 51Crdata showed that the second unlabeled red blood cell units contained from 1.8 to 6.3 per cent (Mean 5.0%) of cells from the first radioactive unit, as carryover. There was no evidence that this led to increased hemoglobin amounts in the supernatant fluids, or diminished in vitro yield or the 24-hour "Cr or Ashby survival. Seam breaks in the IBM centrifuge pack were rare, being as infrequent in single as in double units. All bacteriologic cultures were sterile. All finally reconstituted red blood cell preparations studied by multiple photographs of fields on the scanning electronmicroscope were uniform (Fig. 5). These contained innumerable normal discocytes without membrane damage.

lein pack also are not directly exposed to the outer layer of the freezing container, so in such large volume units less hemolytic damage is expected. The time-temperature curve indicates more rapid thawing than other methods %RECOVERY AS Hb MASS X

'90 0°1 .~ 00 70

60

50 40

30 20 10 l

Discussion

There was no indication of cryogenic red blood cell damage during the rapid physical change to -196 C . Aluminum offers faster thermal conductivity and freezing then do other metals.12*20 The cells outside the polyo-

INDEX OF T t i E R A P E UT IC VIABILITY

-

Al.-Poly (IBMI

A I, Pol y

90.1%

90.6%

LIBM)

I

DOUBLE N.30ur

90.I % a 93.5% S 84.2%

90.6a 92.7 83.9

FIG.4. Final red blood cell recoveries, expressed as hemoglobin mass, including 15 doubly reconstituted units and 12 single ones. The index of therapeutic viability for both methods is given.

Volume 16 Number 2

119

LOW-GLYCEROL CRYOPRESERVED RED BLOOD CELLS Table 1. Metabolic Studies on Modified Low-Glycerol Cryopreserved Red Blood Cells 2, 3-DPG

Unit

Glucose Utilization

Initial

Thaw

Final

ATP

PH

LN-72 LN-73 LN-74 LN-75 LN-103 LN-104 LN-115 LN-165

1.78 1.53 1.79 3.1 1.59 1.68 1.55 1.65

3,900 4,305 4,300 4,600 5,058 4,761 4,540 4.500

2,575 3,585 2,885 2.360 3,565 4.095 3,550 4,360

1,495 3,555 2,365 2,350 3,147 3,508 3,385 4.300

0.94 1.o 1.09 1.12 1.1 1.o 1.2 0.97

6.95 7.1 7.0 6.9 7.0 6.9 7.0 7 .O

1.2

7.35

Normal

1.7

* 0.2 umlml rbclhr

d e ~ c r i b e d . ' ~ *The ~ ' ~operator ~' can visualize the cells as early as one minute after initiating this process, so that the total thawing time can be manually improved, minimizing postthaw hemolysis. Seam leaks that were often self-sealing appear in the polyolein packs during thawing (incidence

4400u (4.4)

about 3 per cent). This permitted any liquid nitrogen that had entered to convert to its vapor phase causing the unit to bulge. These units have been sterile thus far when cultured but must be discarded since liquid nitrogen is not assured sterile. The 51Cr 24-hour survival of sequential

FIG.5. Typical scanning electronmicroscope photomicrograph of final red blood cells immediately prior t o their transfusion. (850 x )

120

Transfusion Mar.-Apr. 1976

BOWMAN, ET AL Table 2. Cost Estimates for Freezinqand Thawina Red Blood Cells Glycerolization and Freezing Glycerol-mannitol-saline solution

Expendables, software Liquid nitrogen Technician time at 1 0 0 per cent efficiency ($4.00/hr.) Total

$0.75 $6.30 $0.20 $0.65 __ $7.90

$ 7.90

$3.75 $1.75 $1.85 $1 .oo $0.45 $8.80

$ 8.80

Thawing and Deglycerolization

IBM 2991 software 3.2 per cent sodium chloride solution Buffered 0.9 per cent sodium chloride-dextrose solution Transfer pack Technician time a t 100 per cent efficiency Total Total approximate cost each unit, 1 IBM harness per 2 units

double or single units was most satisfactory at 93.5 and 92.7 per cent. This is equivalent to that of liquid anticoagulated blood stored at 4 C for two days and tran~fused.'~ Ashby survival studies confirmed the excellent cell life of these reconstituted red blood cells. The high in vitro recovery after freezethaw can be attributed not only to the improved thawing techniques, but also to the polyolein packs. These appear to induce minimal hemolysis compared with vaned freezing bags.' The low leukocyte levels in final units aid in preventing HL-A sensitization and have advantages for transplantation candidates and patients with hematologic disease who require frequent transfusions. The lowered glucose utilization and 2,3DPG values may mean that a minor defect in hexokinase activity exists in the intrakellular anerobic metabolism. This would not significantly affect respiratory function of the transfused final red blood cell product. Derrick, et aL5 also noted a minimal increased sodium influx of such cells. The ability to safely use the IBM 2991 software to process two red blood cell units for the same patient is a new adaptation and cost advantage. This final cost analysis compares favorably with others.2' Such an adaptation is not feasible for processing high-glycerol red blood cells, where all

$1 6.70

sterile lines in the harness must be used in deglycerolization.I0 Quality control studies justify this potential use of the IBM 2991 equipment. The excellent morphology by scanning electronmicroscopy of all reconstituted cells attained by this method contrasts sharply with those red blood cells frozen and reconstituted after seven days liquid storage, where scattered echinocytes and crenated forms are visible.28 It suggests that this frozen red blood cell resource is a superior transfusion component.

Acknowledgment We are grateful for the criticism, advice, and opportunity to work in the laboratories of Dr. Arthur Rowe of the New York Blood Center and of Dr. C. Robert Valeri, Naval Research Blood Center, Boston. We wish to thank Katherine Young, MT (ASCP), Suzanne Ritter, MT (ASCP), and Linda Landis, CLA (ASCP) for skilled technical assistance.

References 1.

Bessis, M.:Living Blood Cells and their UltraStructure, New York, Springer-Verlag, 1973, p.

2.

Crosby, W. H., and F. W. Furth: A modification of the benzidine method for measurement of hemoglobin in plasma and urine. Blood 11:380,

724.

1956.

3. Croxton, F. E.: Elementary Statistics with Applications in Medicine. New York, prentice Hall, 1953.

Numkr 16 Volume 2

4.

5.

6.

7.

8. 9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

LOW-GLYCEROL CRYOPRESERVED RED BLOOD CELLS

Davidsohn, T., and J. B. Henry: Todd-Sanford Clinical Diagnosis by Laboratory Methods. Philadelphia, W. B. Saunders Co., 1974. Derrick, J. B., M. Lind; and A. W. Rowe: Studies of the metabolic integrity of human red blood cells after cryopreservation. Transfusion 9:317, 1969. Funder, J., and J. 0. Wieth: Determination of sodium, potassium, and water in human red blood cells. Scand. J. Clin. Lab. Invest. 18:151, 1966. Hornblower, M., and H. T. Meryman: Influence of the container material on the hemolysis of glycerolized red cells after freezing and thawing. Cryobiology 11:317, 1974. Huggins, C. E.: Frozen blood: principles of practical oreservation. Monogr. Surg. Sci. 3:133, 1966. Gibson, J. G., 11, and W. A. Scheitlin: A method employing radioactive chromium for assaying the viability of human erythrocytes. J. Lab. Clin. Med. 46:679, 1955. Gilcher, R.: Personal Communication. Krimsky, I.: D-2,3 Diphosphoglycerate. In Methods of enzymatic analysis, H. U. Bergmeyer, Ed. New York, Academic Press, 1963, p. 238. Krijnen, H. W., A. C. J. Kuivenhoven, and J. J. F. M. DeWit: The preservation of blood cells in the frozen state. Experiences and current methods in the Netherlands. Cryobiology 5:136, 1968. Meryman, H. T., and M. Hornblower: A method for freezing and washing red blood cells using a high glycerol concentration. Transfusion 12:145, 1972. Mohn, J . F., R. M. Lambert, H. S. Bowman, and F. W. Brason: Experimental transfusion of donor plasma containing blood-group antibodies into incompatible normal human recipients. Br. J. Haematol. 7:112, 1961. Mollison, P. L.: Blood Transfusion in Clinical Medicine, 5th ed. Oxford, Blackwell Scientific Publications, 1972, p. 20. Oski, F. A., C. Smith, and E. Brigandi: Red cell metabolism in the premature infant. Apparent inappropriate glucose consumption for cell age. Pediatrics 41:473, 1968. -, E. Brigandi, and L. Noble: Red cell metabolism in the newborn infant. V. Glycolytic intermediates and glycolytic enzymes. Pediatrics 44:84, 1969. Pert, J. H., P. K. Schork, and R. Moore:.A new method of low-temperature blood preservation using liquid nitrogen and a glycerol-sucrose i d ditive. Clin. Research 11:197, 1963. Rowe, A. W., E. Eyster, F. H. Allen, Jr.,'and A. Kellner: Freezing of erythrocytes for trans-

20.

21.

22.

23.

24.

25.

26.

27.

28.

121

fusion by a glycerol-liquid nitrogen procedure. Transfusion 6 5 2 1 , 1966. -, E. Eyster, A. Kellner: Liquid nitrogen preservation of red blood cells for transfusion. A low glycerol-rapid freeze procedure. Cryobiology 5:i 19, 1968. -: Preservation of blood by the low glycerolrapid freeze process. In Red Cell Freezing, Washington, D. C., American Association Blood Banks, 1973, p. 55. Runck, A. H.,and C. R. Valeri: Therapeutic effectiveness of homologous erythrocyte transfusions following frozen storage at -80 C for up to seven years. Transfusion 10:102, 1970. -, and C. R. Valeri: Recovery ofglycerolized red blood cells frozen in liquid nitrogen. Transfusion 9:297, 1969. Smith, A. U.: Prevention of hemolysis during freezing and thawing of red blood cells. Lancet 2:910, 1950. Valeri, C. R.: Cryopreservation and red cell function. In Progress in Transfusion and Transplantation. Washington, D. C . American Association of Blood Banks, 1972, p. 343. -: Principles of cryobiology. High glycerol storage at -80 C and low glycerol storage at 150 C. In Red Cell Freezing, Washington, D. C., American Association of Blood Banks, 1973, p. 3. -: Simplification of the methods for adding and removing glycerol during freezepreservation of human red blood cells with the high o r low glycerol method. Transfusion 15:195, 1975. Weed, R. L., P. L. LaCelle, and M. Udkow: Structure and function of the red cell membrane: changes during storage. In The Human Red Cell In Vitro, T. L. Greenwalt and G. A. Jamieson, Eds. New York, Grune and Stratton, 1973, p. 65.

Herbert S. Bowman, M.D., Professor of Medicine (Hematology), Milton S. Hershey Medical Center; Hematologist, Harrisburg Hospital, Harrisburg, Pennsylvania 17101. Frank A. Oski, M.D., Professor of Pediatrics, University of Syracuse Medical School, Upstate Medical Center, Syracuse, New York. Judith K. Reihart, SBB (ASCP), Special Imrnunohematology Technologist, Harrisburg Hospital. Mary A. Simmonds, M.D., Resident in Medicine, Geisinger Medical Center, Danville, Pa. Roger K. Cunningham, Ph.D., Assistant Professor, Center for Immunology, State University of New Ydrk at Buffalo, Buffalo, New York.