ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 17, No. 5Copyright © 1987, Institute for Clinical Science, Inc.

The Affinity Glvcated Hem oglobin in a Fam ily with H ereditary Spherocytosis and in O ther Non-hemoglobinopathic H em olytic Anemias*

JONATHAN S. KRAUSS, M .D ., DOROTHY A. HAHN, M .D .,

DEBRA HARPER, MT(ASCP), SHEILA SH ELL, MT(ASCP)SH,

and CHARLES R. BAISDEN, M .D .

Section o f Clinical Pathology, Department o f Pathology,

Medical College o f Georgia Hospital and Clinics, Augusta, GA 30912-0300

ABSTRACT

The glycated hemoglobin (GHb) is lowered by hemolytic anemia. The cation-exchange HbAj has been shown to be lowered by hereditary sphe­rocytosis (HS). The HbAj, however, can be increased by elevations of fetal hemoglobin (HbF). The affinity GHb, a param eter related to, bu t not identical with, the HbAx, and unaffected by H bF, has been shown to be low in hemoglobinopathies bu t not, to our knowledge, in HS and other non-hem oglobinopathic hemolytic anemias. Therefore, the affinity G H b and H bF was determ ined in four m em bers of an HS family and in nine o ther cases of non-hemoglobinopathic hemolytic anemia, including th ree autoim m une hemolytic anemias, four red cell fragmentation syndromes (two “W aring b lender” syndrom es, one throm botic throm bocytopenic purpura in association with tumor, and one case of dissem inated intravas- cular coagulation), and two red cell m em brane defects: paroxysmal noc­turnal hemoglobinuria and another case of hereditary spherocytosis. The G H b for these nine cases was 3.6 ± 1.7 percent (normal 6.0 ± 2.0 pe r­cent; p < 0.001). The reticulocyte count, available in four cases, was 0.23 ± 0.14 and correlated negatively with the GHb. The average G H b in the HS family was 3.9 ± 0.8 percent, which was significantly less than the normal of 6.0 ± 2.0 percent (p < 0.001); the H bF was < 1 .0 percent. It is concluded that the G H b is dim inished in hemolytic anemias not asso­ciated with hemoglobinopathies and that this lowering reflects the short­ened red cell life span in these processes. To our knowledge, this is the first report of low GH b in hemolytic anemia not associated with hemoglo­binopathy, by the affinity chromatographic technique, as opposed to the cation-exchange chromatographic technique.

* Address reprint requests to Jonathan S. Krauss, Clinics, BIH 222B, 1120 15th Street, Augusta, GAM .D ., Section of Clinical Pathology, Department of 30912-0300.Pathology, Medical College of Georgia Hospital and

3310091-7370/87/09-0331 $01.20 © Institute for Clinical Science, Inc.

3 3 2 KRAUSS, HAHN, HARPER, SHELL, AND BAISDEN

Introduction

The glycated (glycosylated) hemoglo­bin (GHb, HbA1( HbAlc) reflects irre ­versible, predom inantly non-enzymatic glycation of the N H 2-term inal amino acid of the beta-chain of hemoglobin by the keto-amine linkage. The G H b is formed slowly and continuously throughout the 120 day red blood cell (rbc) life-span and w ill r e f l e c t rb c a g e .7 T h e c a tio n - exchange HbAlc has been shown to be d ec reased by various hem oly tic a n e ­mias,6 the GH b, as determ ined by affin­ity chrom atography, a re la ted b u t not equivalent procedure, has been shown to be lower in various hem oglobinop­athies1,12 but not, to our knowledge, in o th e r hem oly tic anem ias. A study is reported of the affinity G H b proportion in nine patients with hemolytic anemias, including one with hereditary spherocy­tosis (HS) not known to be associated with hemoglobin abnormalities. H eredi­tary spherocytosis (HS) is the most com­m on hered ita ry hem olytic d iso rder in N orthern Europeans, an autosomal dom ­inant condition of lim ited penetrance in which a defect can exist in spectrin-pro- tein-4.1 binding.2’3,14 Panzer et al stud­ied two patients with HS and showed the cation-exchange HbA2 to be decreased by th is p ro c e s s .21 T he H bA x can be falsely e lev a ted by fetal hem oglob in (HbF) which can be increased by hem o­lysis.9 The affinity G H b has never been m easured in HS, however, and is unaf­fected by HbF. Therefore, five m em bers of a family with HS w ere studied and their affinity G H b and H bF levels de ter­m in ed , as w ell as th e H bA j in one patient.

Methods

The rbc indices and H b were de te r­m ined by a C oulter S + V blood cell counter.* Reticulocyte counts were per­

form ed by a supra-vital staining tech­nique employing new m ethylene b lue.10 The GH b was perform ed by Abraham’s a f f i n i t y m i c r o c h r o m a t o g r a p h i c m ethod, f 1,12,23 H ere e thy lened iam ine tetraacetic acid (EDTA) anti-coagulated whole blood stored at 4°C is hemolyzed and applied to the micro-column. In the first step, the non-glycated hemoglobin is e lu ted by the buffer. In the second step , the G H b is com petitively e lu ted from the phenyl-boronic acid agarose by so rb ito l. B lood glucoses w ere d e te r ­m ined by the hexokinase m ethod on the SM A -II.$ The H b F was m easured by either alkali-denaturationf4 or by radial im m uno-diffusion .§18 HbA j was m ea­sured by a cation-exchange m icrochro­matographic p rocedu re!15 with a labile hemoglobin eliminator. A red cell agglu­tination m ethod was used to perform the direct antiglobulin test (Coombs’, differ­ential agglutination titer [DAT]).

Case Reports (See tables I and II)(A) Autoimmune Hemolytic Anemia (AIHA)Patients # 1 , 2, and 3 had long standing Coombs

positive anemias with reticulocytosis. Patient # 1 had systemic lupus erythematosus (SLE) and was given corticosteroids; she returned three months later with an increased hemoglobin and a GHb of 6.6 percent.

(B) Red Cell Fragmentation SyndromesPatients # 4 and 5 had mitral and aortic valvar

replacement respectively (MVR, AVR). Patient # 6 had laryngeal cancer and terminally developed schis- tocytosis and thrombocytopenia with normal clotting studies suggestive of thrombotic thrombocytopenic purpura (TTP). Patient # 7 had a craniopharyngioma and developed sepsis with disseminated intravascu- lar coagulation (DIC) postoperatively and expired. All three patients had at least moderate schistocytosis on blood smear (1 to 3 schistocytes per oil immersion field) (see figure 1, patient # 4 ). All blood smears were reviewed by a pathologist.

(C) Membrane DefectsPatient # 8 was an 18-year-old girl with pancyto­

penia, hypocellular marrow, macrocytosis, 34 per­cent fetal hemoglobin (HbF), a positive Ham’s and sucrose lysis test, and a decreased leukocyte alkaline phosphatase (LAP) score of 19, all of which are con­sistent with paroxysmal nocturnal hemoglobinuria

* Coulter Electronics Inc., Hialeah, FL.

t Isolab Inc., Akron, OH.$ Technicon Instruments Corp., Tarrytown, NY. § H elena Labs, Beaumont, TX.

GLYCATED H E M O G L O B IN IN H E R E D IT Y SPH ER O C Y TO SIS 333TABLE I

Patient Data

Patient Diagnosis Sex AgeGHb

(percent) Retie Glucose (nrnol/L)

1 AIHA F 23 2.7 0.200 5.02 A IHA M 8 0.8 0.320 6.73 AIHA F 58 3.0 - 4.94 A VR M 55 4.9 - 8.65 M V R F 33 3.5 - 4.76 Ca/TTP M 64 5.5 - 10.77 DIC M 64 6.2 - 12.68 PNH F 18 3.5 0.400 6.39 HS M 4 2.6 0.340 -

N = 9 36 ± 24 3.6* ± 1.7 0.23* ± 0.14 7.4 ± 2.9NormalRange: (6.0 ± 2.0) (0.001 - 0.024) 3.9 - 6.1 (fasting)

*P < 0.001 C a / T T P = c a n c e r c u m t h r o m b o t i c t h r o m b o c y t o p e n i c p u r p u r aGHb = glyca t e d hemoglobin DIC = disseminated intravascular coagulationAIHA = autoimmune hemolytic anemia P NH = p a roxysmal nocturnal hemoglobinuriaA V R = aortic v a lvar replacement HS = h e reditary spherocytosisM V R = mitral valvar replacement

(PNH). Patient # 9 was a 4-year-old Caucasian boy w hose m other and m aternal grandm other w ere known to have hereditary spherocytosis (HS). His osmotic fragility was increased, and his blood smear had spherocytes (see figure 1, patient #9).

Sickled red blood cells were not identified on any blood smears from these nine patients, six o f whom w ere black. This study em ployed a retrospective review of these patients’ laboratory data and charts. None of the patients was known to be diabetic. Ran­dom blood glucoses were included (table I).

(D) Family with Hereditary Spherocytosis Three non-diabetic caucasian patients (a brother,

sister, and father, DJB, BB, and RB, aged 10, 8, and34, respectively) underwent splenectom y for a long­standing hemolytic process associated with sphero­cytes on blood smear (figure 2). All three spleens were enlarged and had congested sinusoids with rela­tively little pigm ent (figure 3). Two-thirds of the patients had cholecystectomy for pigm ent stones as well. Other patients examined included a sister (LB, aged 11) and a paternal aunt (WBB).

TABLE I I

Family Laboratory Data

Kb (g/L) MCHC (g/L)GHb

(percent) (percent)H bF

(percent)

BBDJBRBLBWBB

Normal Range :

94103103136132

379359359347361

0.0800.0890.043

0.092

0.076 ± 0.022*

0.001 ± 0.024

3.53.73.05.14.1

6.0 ± 2.0

5.2*

7.0 ± 1.5

<1.0<1.0<1.0<1.0<1.0<1.0

*p < 0.001 Hb - hemoglobinM CHC = m e a n cell hemoglobin concentration GHb = glyca t e d hemoglobin H bA = adult h e m oglobin HbF = fetal hemoglobin

334 KRAUSS, HAHN, HARPER, SHELL, AND BAISDEN

F i g u r e 1 . B l o o d sm ears from patient # 4 (top) and patient # 9 (bot­tom) showing erythrocytic fragmentation and sphero- cytes respectively (origi­nal magnification 1,OOOX; W right’s stain).

Results

For the first nine patients, the GHb was 3.6 ± 1.7 percent (normal: 6.0 ±2.0 percent); p < 0.001 by the standard e rro r of th e d ifference be tw een two means. Reticulocyte counts were avail­able in four patients (#1, 2, 8, and 9) and w ere 0 .23 ± 0 .14 (norm al 0 .001 to0 .0 2 4 ). Two o f th e s e p a t ie n ts had autoimm une hemolytic anemia (AIHA), one had HS, and the o ther had PNH. T heir average age was 13 ± 9 years. Their GH b was 2.4 ± 1 .1 (p < 0.001).

The equation y = mx -I- b where y is the reticulocyte count and x the GH b for this group of patients was y = 4.5x + 43 percent; the correlation coefficient, r, was —0.72. The random blood glucose in eight patien ts was 7.4 ± 2 . 9 mmol per L.

Four m em bers of the HS family tested w ere show n to have reticu locy tosis, which averaged 7.6 percent. Four of the five had spherocytes on the blood smear (figure 2). The m ean cell hem oglobin concentration (MCHC) was elevated to 360 g per L and the G H b was dimin-

GLYCATED HEMOGLOBIN IN HEREDITY SPHEROCYTOSIS 3 3 5

F i g u r e 2. N o te th e spherocvtes on the blood smear (BB, Wright’s stain, o r ig in a l m a g n ifica tio n 1,250 x ).

o

Wo i

ished to 3.9 percent (table II). None of the kindred had elevations of HbF. The father, RB, had a low HbAj of 5.2 per­cent. A sister, LB, had a Hb of 136 g per L, M CH C of 347 g p e r L, an affinity G H b of 5.1 percent, and an H bF of <1 p e r c e n t . T h e d i r e c t a n t ig lo b u l in (Coombs’) test was negative in all tested.

Discussion

Long-lived proteins are glycated at the am ino-term inus, form Schiff-bases, are co n v erted to A m adori p roducts , and th e n a re ir r e v e r s ib ly c o n v e r te d to

advanced glycation end products.5 These products will accum ulate throughout the lifetime of the red cell and the GH b will be approximated by the equation y = kt where y is the G H b, k is a constant, and t is the life-time of the red cell.6 Thus, th e g ly ca ted h e m o g lo b in le v e l w ill re f le c t c h ro n ic b lood g lucose levels when the red blood cells have a normal life span. H ence, this m easurem ent is useful in the treatm ent of diabetes mel- litus.716 Conversely, if the red cell life span is shortened, as it is in hemolytic anemia, and the blood glucose level is within the range of normal, as it can be

336 KRAUSS, HAHN, HARPER, SHELL, AND BAISDEN

t 9 «r .rw1

♦

v ? ,' • ' V * t / - y * * »

£ £ % * > * > • *

STp **, h iK i . U '

F i g u r e 3 . S p l e e n from patient DJB. N ote th e c o n g e s t io n o f th e cords o f Billroth and the lack of pigm ent (original m a g n if ic a t io n 5 0 0 x , H&E).

in the non-diabetic, the low G H b should r e f le c t th e s h o r te n e d re d ce ll life . Patient # 1 is an excellent example of the G H b as a m easure of rbc age. D uring acute hemolysis, her G H b was 2.7 per­cent with 20 percent reticulocytes; three m onths later, after the hem olysis had abated, her G H b was 6.6 percent.

This hypothesis is substan tia ted by studying three patients with AIHA and four patients with red cell fragmentation syndrom es,19 as characterized by schis- tocytosis in two p a tien ts w ith valvar lesions w here rbc destruction occurs in

the heart and two patients with throm ­bo tic th ro m b o cy to p en ia p u rp u ra and D IC , respectively, w here the red cell destruction occurs in the small vessels, and in two patients with rbc m em brane defects. In PNH, the m em brane is espe­cially sensitive to com plem ent-m ediated lysis11 and aplastic anemia develops in 25 percent of the patients, as it did here;8 w hereas in HS, the defect can be related to spectrin-protein-4.1 binding.2,3,14

Although the cation-exchange HbAlc has b e e n show n to be d e c re a sed in undefined hem olytic anem ias7, to our

GLYCATED HEMOGLOBIN IN HEREDITY SPHEROCYTOSIS 3 3 7

knowledge the affinity G H b has not been d e te rm in ed in hem olytic anem ia—not necessarily owing to hemoglobinopathy. O nly abou t 70 p e rc e n t of th e H bA lc recovered from ion-exchange chromatog­raphy has been shown to be glycated by affin ity c h ro m a to g rap h y ; h e n c e , th e m ethods do not m easure the same spe­cies.20 Therefore, despite the good cor­rela tion b e tw een the two m ethods, a study such as this is justified.

Patients # 4 to # 7 with rbc fragmenta­tion syndromes had an average GH b of5 .0 ± 1 .1 p e r c e n t w h e r e a s th e rem ainder of the patients with AIHA and rbc m em brane defects had an average G H b of 2 .5 ± 1.0 p e rc e n t. Random blood glucose for the form er was 9.2 ±3.3 m mol per L, as opposed to 5.7 ±0.83 mmol per L for the latter patients, which was suggestive of abnorm al glu­cose tolerance in the form er, possibly related to “stress.” All four were hospi­talized and seriously ill; patients # 6 and # 7 later expired. These patients w ere in c lu d ed , how ever, because of th e ir m arked schistocytosis, the low normality of their GHbs, and their clinical picture. Possibly th e G H b is h ig h er and less helpful in rbc fragm entation syndromes owing to transfusion,22,24 low er rate of h em o ly s is , o r a p ro lo n g e d h o sp ita l course with glucose intolerance.

A lthough six of these n ine patien ts were black, none had sickled RBCs on blood smear, thus diminishing the possi­bility of an associated sickling hemoglo­binopathy. It is believed by us that the negative correlation of the G H b with the reticulocyte count suggests the GH b can be used to judge the red cell life in esti­mating the severity of hemolysis as well as to differentiate hemolytic from non­hem olytic anem ias in norm oglycem ic patients because the reticulocyte count is inversely related to the rbc age.

Therefore, the G H b is recom m ended by the p resen t authors as an ancillary te s t to th e re ticu lo cy te coun t in the workup of hem olytic anemia, w here it

could be a low-cost substitute for radio­isotopic ex vivo stud ies and w here it would represen t a more chronic m easure of hemolysis.

Finally, it is believed that the family studied has HS owing to (1) family data- autosomal dom inant inheritance in this caucasian family; hematologic data, (2) rbc in d ic e s -e le v a te d M C H C (HS is one o f th e few causes o f an M C H C g re a te r th a n 3613) (3) cy to log ic e v i­dence-spherocy tes on the blood sm ear (figure 2), (4) reticulocytosis compatible with hemolysis, (5) histologic ev idence- enlarged spleens with congestion of the cords of B illroth and a relative lack of p igm en t, characteris tic of HS (figure 3)17, and (6) DAT negativity. Moreover, the sister (LB) was most likely unaffected by HS as was indicated by her laboratory data.

Panzer e t al determ ined the HbAi in two HS patients by the cation-exchange m ethod .21 Because the results by this m ethod can be elevated by H b F ,12,16,20 it was decided to m easure H bF in this fam­ily as well. M oreover, there is a report of one patient with elevated H bF in H S .9 O ur data are consistent with lack of ele­vation of H b F in HS; H bF elevation with hemolysis is much m ore common in hemoglobinopathies.

O ur data are compatible with a low­ered affinity G H b in HS. It is our belief that this lowered GH b reflects the short­ened rbc life associated with HS. The normal H bF levels in this family suggest th a t th e cation-exchange HbAx in HS sh o u ld b e a c cu ra te . T h e re fo re , o u r beliefs agree w ith Panzer in tha t the HbAj should reflect the shortened rbc life in HS. The HbA]^ was perform ed on RB and was found to be low, as well, at 5.2 percen t. Samples from 20 diabetic patients w ere examined in our laboratory and a c o rre la tio n coeffic ien t, r, was obtained of 0.95 for the GH b and HbAx; hence, agreem ent should be good. Thus, e ither the affinity G H b or the cation- exchange HbAx can be used to m easure

3 3 8 KRAUSS, HAHN, HARPER, SHELL, AND BAISDEN

rbc life in HS. However, because HS can p resent in infancy, when H bF levels are higher, it is our be lief that the affinity G H b may be preferable to the cation- exchange HbAj, particularly in patients less than one year old.

The average G H b here of 4.1 percent corresponds to an HbAx of 4.5 p ercen t,1 which represents a 51Cr-tagged rbc sur­vival of 12.5 to 15 days w here the normal ranged from 30 to 80 days.21 Hence, the low G H b here reflects the substantially shortened HS rbc life of less than one- half normal.

It is concluded that the relationship of the G H b and the reticulocyte count to rbc age in hemolytic anemia is similar to the re la tio n sh ip of th e G H b and the blood glucose to glucose homeostasis in diabetes. In hemolytic anemia, the GHb d irectly reflects RBC age; w hereas in d iabetes, the G H b reflec ts long term blood glucose levels.

References

1. A b r a h a m , E . C . , P e r r y , R . E . , and St a l l in g s , M.: Application of affinity chromatography for separation and quantitation o f glycosylated hemoglobins. J. Lab. Clin. Med. 102:187-97,1983.

2. A g r e , P., O r r in g e r , E. P., and B e n n e t t , V .: Deficient red-cell spectrin in severe, recessively inherited spherocytosis. N ew Engl. J. M ed. 306:1155-1161, 1982.

3. A g r e , P., A s i m o s , A ., C a s e l l a , J. F ., and M c M i l l a n , C .: Inheritance pattern and clinical response to splenectom y as a reflection of eryth­rocyte spectrin deficiency in hereditary sphero­cytosis. N ew Engl. J. M ed. 3 i5 :1 5 7 9 -1 5 8 3 , 1986.

4. B e t k e , K., M a r t i , H . Q . , and Sc h l i c h t , I.: Estimation of small percentages o f foetal haemo­globin. Nature (London) 184:1877-1878, 1959.

5. B r o w n l e e , M ., V l a ssa r a , H . , and C e r a m i , A.: Nonenzymatic glycosylation and the pathogene­sis o f diabetic complications. Ann. Intern. Med. 101:527 - 537, 1984.

6. B u n n , H . F., H a n e y , D . N ., K a m in , S., and G a b b a y , K. H .: The biosynthesis o f human hemoglobin Alc. Slow glycosylation of hemoglo­bin in vivo. J. Clin. Invest. 57:1652-1659, 1976.

7. B u n n , H . F ., G a b b a y , K . H . , and G a l l o p , P. M.: The glycosylation o f hemoglobin: rele­vance to diabetes mellitus. Science 200:2 1 -2 7 , 1978.

8. C a m it t a , B . M ., S t o r b , R . , and T h o m a s ,E. D.: Aplastic anemia: Pathogenesis, diagnosis,

treatment, and prognosis. New Engl. J. Med. 306:6 4 5 -6 5 2 , 1982.

9. C o o p e r , H . A. and H o a g l a n d , H . C.: Fetal hem oglobin . Mayo Clin. Proc. 4 7 :4 0 2 -4 1 4 , 1972.

10. D a c i e , J. V. and L e w is , S. M.: Practical Hem a­tology, 6th ed. New York, Churchill Livingstone,1984. pp. 5 8 -6 1 .

11. D e s s y p r is , E. N ., C l a r k , D . A ., M c K e e , L. C ., and K r a n t z , S. B.: Increased sensitivity to com plem ent o f erythroid and myeloid pro­genitors in paroxysmal nocturnal hem oglobin­uria. N ew Engl. J. Med. 309:690-693, 1983.

12. F a ir b a n k s , V. F. and Z im m e r m a n , B. R.: Mea­surement of glycosylated hemoglobin by affinity chromatography. Mayo Clin. Proc. 58:770-773,1983.

13. F i s c h e r , S. L. and F is c h e r , S. P.: Mean cor­puscular volume. Arch. Intern. Med. 143:2 8 2 - 283, 1983.

14. G o o d m a n , S . R., S h i f f e r , K. A ., C a s o r ia , L. A., and E y s t e r , M. E .: Identification of the molecular defect in the erythrocyte membrane sk ele to n o f som e kindreds w ith hered itary spherocytosis. Blood 60:772-84, 1982.

15. H a m m o n s , G. T ., J u n g e r , K ., M c D o n a l d , J. M ., and L a d e n s o n , J. H .: Evaluation of three minicolumn procedures for measuring hem oglo­bin Aj. Clin. Chem. 28:1775-1778, 1982.

16. H e a l t h and P u b l ic P o l ic y C o m m i t t e e , A C P. Glycosylated hemoglobin assays in the manage­ment and diagnosis o f diabetes mellitus. Ann. Intern. Med. 101:710-713 , 1984.

17. K is s a n e , J. M.: Anderson’s Pathology. St. Louis,C. V Mosby Co., 1985, pp. 1257-1351.

18. K o h n , J. a n d PAYNE, B. V.: A n im m u n o lo g ic a l m e th o d fo r t h e d e te c t io n a n d e s t im a tio n o f fe ta l h e m o g lo b in . J. C lin . P a th o l. 2 5 : 8 3 0 - 8 3 1 , 1972.

19. N a n d , S . , B a n s a l , V. K ., K o z e n y , G . , V e r - t u n o , L ., e t a l.: Red cell fragmentation syn­drome with the use o f subclavian hemodialysis catheters. Arch. Intern. Med. i4 o : 1 4 2 1 - 1 4 2 3 ,1985.

20. L a d e n s o n , J. H ., C h a n , K.-M ., and K il z e r , P.: Glycated hemoglobin and diabetes: A case and an o v e r v ie w o f th e su b jec t. C lin . C h em . 31:1060-1067 , 1985.

21. Pa n z e r , S., K r o n ik , G., L e c h n e r , K ., B e t t e l - HEIM , P ., e t al.: G lycosylated hem oglobins (GHb): An index of red cell survival. B lood 59:1348-1350, 1982.

22. W e in b l a t t , M. E ., K o c h e n , J. A., and Sci- MECA, P. G .: Chronically transfused patients with increased hem oglobin AIC secondary to donor blood. Ann. Clin. Lab. Sci. i6 :3 4 -3 7 ,1986.

23 . Ya t s c o f f , R. W ., T e v a a r w e r k , G . J. M ., C la r - s o n , C. L., and W a r n o c k , L. M.: Evaluation of an affinity chromatographic procedure for the d eterm in a tio n o f g ly cosy la ted hem oglob in (HbAx) Clin. Biochem. 7 6 :2 9 1 - 2 9 5 , 1983 .

24 . Z e l l e r , W. P ., E y z a g u ir r e , M ., H a n n ig a n , J., O z o g , K ., e t al.: Fast hem oglobins and red blood cell metabolites in citrate phosphate dex­trose adenine stored blood. Ann. Clin. Lab. Sci. i5 :6 1 -6 5 , 1985.