Ann Hematol (1992) 64:88-92
Annals of
Hematology 9 Springer-Verlag 1992
Original article Prevalence of increased osmotic fragility of erythrocytes in German blood donors: screening using a modified glycerol lysis test S.W. Eber, A. Pekrun, A. Neufeldt, and W. Schr6ter Department of Pediatrics, University of G6ttingen, Robert-Koch-Strasse40, W-3400 G6ttingen, Federal Republic of Germany Received August 20, 1991/Accepted November 27, 1991
Summary. We screened for increased osmotic fragility of erythrocytes in 1464 healthy German blood donors. The osmotic fragility was determined by an acidified glycerol lysis test (AGLT) using glycerol-sodium phosphate-buffered NaC1 solution. Since the original test described by Zanella et al. [23] showed only low specificity for hereditary spherocytosis, we used a modification with 0.0093 M sodium phosphate-buffered glycerol-saline solution, pH 6.90, instead of the original 0.0053 M sodium phosphate buffer, pH 6.85. Sixteen of the donors (1.1%) had a "pathologic result," similar to that of 32 patients with hereditary spherocytosis: AGLT 50 < 5 rain ("half-time of AGLT, defining normal and pathologic results). The osmotic fragility of the erythrocytes from 12 of these donors was further investigated using the conventional test with hypotonic NaC1 solutions. With one exception, increased osmotic fragility was verified in all of them by both tests. Further hematologic data showed a mild reticulocytosis (2% and 2.6%) in two of the donors. One donor had a moderate reticulocytosis of 6.5%, probably due to a mild, previously undiagnosed spherocytosis; 99 of the donors had an intermediate result (AGLT 50: 5 - 30 min). Hypotonic lysis of their erythrocytes by the conventional method showed a normal result; there were no signs of increased hemolysis. Thus they are not definitely regarded as having increased osmotic fragility of their erythrocytes. Erythrocyte osmotic fragility shows a wide distribution range in the normal population and might be normally distributed. Thus the blood donors with "pathologic AGLT ( < 5 min)" probably represent only one end of a continuum of salt-dependent hemolysis, and not a separate entity. However, they did show additional minor signs of a functional defect of the erythrocyte membrane and therefore could be carriers of a spherocytosis trait. The frequency of carriers of an erythrocyte membrane defect (possible spherocytosis trait) could be as high as 1.1% in the general population and would distinctly exceed the prevalence of patients with apparent spherocytosis (0.02%). Address for correspondence: S.W. Eber
Key words: Osmotic fragility - Acidified glycerol lysis test - Hemolytic anemia - Hereditary spherocytosis
Introduction Hereditary spherocytosis is the most common congenital hemolytic anemia in middle Europe. Its incidence in the Caucasian population was estimated to be 200-300 per million [6, 15]. These figures are probably an underestimate, because about a quarter of the patients show a mild clinical course [12] and therefore the diagnosis of spherocytosis may be missed. Hemolytic anemia ranges from mild, fully compensated forms to severe recessive forms, requiring regular transfusions [9]. In 20% of the patients the manner of inheritance is recessive, with both parents being asymptomatic, heterozygous carriers [2, 3, 9]. A high prevalence of carriers with a symptomless trait for spherocytosis was reported by Dacie [6] and by Godal and Heisto [10]. Spherocytosis seems to be a heterogeneous clinical entity, and primary defects of spectrin [22], ankyrin [14], band 3 [13], and band 4.2 [19] have been described. The determination of the osmotic fragility of the erythrocytes is widely used as a diagnostic tool for spherocytosis. Therefore, it seemed appropriate to use a rapid and sensitive test of osmotic fragility in a population screening for spherocytosis trait. However, measuring the osmotic fragility in hypotonic NaC1 solution [17] is time consuming, and diagnosis fails in many of the mild cases [6]. Gottfried and Robertson [11] used the lysis time of erythrocytes in isotonic glycerol solution, pH 7.4. Because of the relatively low sensitivity of this "glycerol lysis test" Zanella et al. [23] proposed the "acidified glycerol lysis test" (AGLT) using 0.0053 M sodium phosphate-buffered glycerol-saline solution, pH 6.85. They claimed 100% sensitivity for spherocytosis. Bucx et al. [4] found the same sensitivity using 24-h incubated blood instead of fresh blood. However, in a re-evaluation, Rutherford et al. [18] found a considerable overlap between
89 spherocytosis p a t i e n t s a n d controls. We therefore m o d i fied the A G L T by using a higher s o d i u m p h o s p h a t e - b u f fer c o n c e n t r a t i o n a n d a slightly increased p H value. W i t h this m o d i f i c a t i o n b o t h high specificity a n d high sensitivity were achieved, w h i c h allowed for p o p u l a t i o n screening. We tested 1464 n o r m a l b l o o d d o n o r s . T h e o s m o t i c fragility o f erythrocytes f r o m d o n o r s with " p a t h o l o g i c " A G L T results was f u r t h e r investigated with the conventional test, using h y p o t o n i c salt s o l u t i o n s [17]. A n increased o s m o t i c h e m o l y s i s was f o u n d for m o s t o f t h e m . A d d i tionally, the b l o o d d o n o r s were tested for m i n o r signs o f increased hemolysis. It was thus p o s s i b l e to evaluate the f r e q u e n c y o f clinically i n a p p a r e n t carriers for spherocytosis.
Methods
Patients and normal blood donors A total of 1464 blood donors (1127 male) to the blood transfusion center of the University Hospital of G0ttingen were investigated. The donors were healthy and had no signs of infection. Complete blood cell count, liver enzymes, and renal function were normal. Serologic tests for viral hepatitis and lues were negative. Most donors were 20-31 years old (medium age 25 years). In addition, 44 German patients (19 male) with spherocytosis at the Department of Pediatrics, University of GOttingen, and some other hospitals were investigated. The age of the children was 6 months to 18 years. All patients had hemolytic anemia of varying degree; diagnosis was made by the detection of spherocytes in the blood smear and proof of increased osmotic fragility of the erythrocytes. A reduced concentration of spectrin (below 80% of normal) in the erythrocyte membrane was found in nearly all patients with moderate to severe hemolytic anemia (hemoglobin concentration < 120 g/l, reticulocyte count > 6%, bilirubin concentration > 34 amol/1). Patients with mild disease had a compensated hemolytic anemia (hemoglobin concentration 110-150 g/l, reticulocyte count < 6%, bilirubin concentration up to 34 amol/1) and a normal spectrin concentration [9].
5 ml phosphate-buffered saline (PBS; 0.01-0.05 M sodium phosphate buffer, pH 6.90, containing 0.135 MNaC1). The pH value of the sodium phosphate buffer was adjusted to pH 6.90 _+ 0.05 at 25~ by mixing primary and secondary sodium phosphate solutions containing the desired molarity without further addition of NaOH or HC1 in order to avoid a rise of the osmolarity. To 1 ml of this suspension 2 ml of glycerol solution (glycerol 0.3 M, sodium chloride 0.041 M, sodium phosphate buffer 0.003-0.015 M, pH 6.90, corresponding to the different sodium phosphate concentrations in PBS) was added. The final concentrations in the assay were as follows: 0.0053-0.0265 M sodium phosphate, 0.072 M NaC1, 0.2 M glycerol; final osmolarities ranged from 0.356 to 0.406 osmol/1. After the initial evaluation of specificity and sensitivity of the test, all further measurements were done with 0.0093 M sodium phosphate-buffered glycerol-saline solution, pH 6.90 (final osmolarity: 0.366 osmol/1). The rate of hemolysis was followed by measuring the decline of "extinction" at 625 nm, which is due mainly to the turbidity of the suspended erythrocytes. Half-time of the test (AGLT 50) was defined as the time required for the "extinction" to fall to half of the initial value. To avoid sedimentation of the erythrocytes during the 30-rain incubation, the cuvettes were shaken gently every 10 rain.
Results
Sensitivity and specificity of the acidified glycerol lysis test T h e o r i g i n a l test d e s c r i b e d by Z a n e l l a et al. [23] using 0.0053 M p h o s p h a t e - b u f f e r e d glycerol-saline s o l u t i o n gave a c o n s i d e r a b l e overlap b e t w e e n p a t i e n t s with spherocytosis a n d n o r m a l controls (Fig. 1). Five o f the n o r m a l c o n t r o l values fell below 5 rain. Therefore, we c h a n g e d the c o n c e n t r a t i o n o f s o d i u m p h o s p h a t e b u f f e r until b o t h m a x i m a l sensitivity a n d specificity were r e a c h e d (Fig. 2). Using a s o d i u m p h o s p h a t e - b u f f e r c o n c e n t r a t i o n o f 0.0093 M , 32 p a t i e n t s with spherocytosis consistently h a d an A G L T 50 value below 5 m i n (15 s - 4 . 5 min, with a m a x i m u m frequency between 1 a n d 2 min, d a t a n o t shown).
Hematologic data, osmotic fragility Hematologic data were obtained by routine methods according to Dacie and Lewis [7]. Normal hemoglobin concentrations of agematched children were taken [16]; normal reticulocyte range was 0.5%-1.5%, normal total bilirubin concentration less than 17 pmol/1. Spectrin concentration of the erythrocytes was determined by means of an ELISA using monoclonal spectrin antibodies (normal mean _+SD: 226 _+ 35 xl03 copies per cell) [8]. Osmotic fragility of erythrocytes in hypotonic NaC1 solutions was determined in fresh and 24-h-incubated blood using the method of Parpart et al. [171. For a semiquantitative evaluation, the NaC1 concentration (g/dl) was determined at the onset of osmotic hemolysis (1070hemolysis), at the median (50~ hemolysis), and at completion (100~ hemolysis). Normal ranges (two standard deviations) were as previously published [20]. The osmotic fragility in fresh or incubated blood samples of the patients was designated as slightly increased if one or two of the measured values were outside the normal range; if all three values were outside the normal limits, the osmotic fragility was described as distinctly increased. AGLT was done according to the method of Zanella et al. [23] at 25~ hut using different sodium phosphate-buffered glycerolsaline solutions. The pH value was increased to pH 6.9 +_ 0.05. Control of pH as well as of temperature is rather critical: decrease of the temperature to 20~ gave a shortened half-time of the test (minimum 15 min), increase to 37~ a distinctly prolonged halftime. EDTA-anticoagulated whole blood (20 ~1) was mixed with
Screening for increased osmotic fragility of healthy blood donors We tested 1464 h e a l t h y b l o o d d o n o r s . T h e results are given in Table 1. T h e m a j o r i t y o f the d o n o r s h a d n o r m a l A G L T 50 values above 30 min. Sixteen d o n o r s h a d a clear decrease o f A G L T 50 below 5 min, o v e r l a p p i n g with t h a t o f p a t i e n t s with spherocytosis. T h e result o f the A G L T is therefore d e s i g n a t e d as " p a t h o l o g i c " . Twelve o f the 16 d o n o r s were f u r t h e r evaluated. Clinical h i s t o r y was uneventful, giving no hint for a h e m o l y t i c a n e m i a in the f a m i l y history; the spleen was n o t increased. T h e o s m o t i c fragility o f their erythrocytes was tested with h y p o t o n i c NaC1 s o l u t i o n s in fresh a n d i n c u b a t e d b l o o d samples. E i g h t o f these d o n o r s h a d slightly increased o s m o t i c fragility o n l y after 24-h i n c u b a t i o n o f b l o o d samples. T h e i r h e m o g l o b i n c o n c e n t r a t i o n (range 1 3 4 - 1 6 3 g / l ) a n d bilir u b i n c o n c e n t r a t i o n (range 3 . 4 - 1 0 . 2 p m o l / 1 ) were n o r mal; the reticulocyte counts were n o t increased (range 0 . 3 % - 1 % ) . Two o f the d o n o r s with " p a t h o l o g i c " A G L T h a d a slightly increased o s m o t i c fragility b o t h in fresh a n d in i n c u b a t e d b l o o d samples. B o t h h a d n o r m a l h e m o -
90
Normal controls
9oeee 9149
Hereditary spherocytosis
ii
9
eee
9 9149 9
9 1 4 9 1 4 9 1 49 9 1e4e99 1 4 9 1 4 9
~9 9
L 0
I 5
I I I 10 15 20 AGLT 5o (minutes)
I 25
.t/" 30
Fig. 1. AGLT results of 14 patients with spherocytosis and 186 normal children (age 1-18 years) using 0.0053 M phosphate-buffered glycerol-saline solution, pH 6.85. ;~'s~%indicates 155 AGLT 50 values above 30 min "" 9
Specificity (%)
Sensitivity (%)
100
100
I
l
50
0
PN 2,7
Discussion 50
0 PN PN PN PN PN PN PN PN 5~3 6~6 8~0 9~3 10~6 13,3 15,9 26,5 [ram phosph{atebuffer )
Fig. 2. Sensitivity and specificity of the AGLT at various sodium phosphate-buffer concentrations in the glycerol-saline solution (final concentrations): P, Patients (n = 9); N, normal controls (n = 18). Sensitivity (hatched bars) is indicated by the percentage of patients with spherocytosis giving a pathologic result (AGLT 50 value < 5 rain), specificity (blank bars) by the percentage of normals showing a normal result (AGLT 50 value > 30 min)
Table 1. Acidified glycerol lysis test in healthy blood donors a AGLT 50 (min)
Classification
0- 5 5 - 30
Pathologic Possibly pathologic Normal
> 30 Total
globin concentrations but disclosed a mild reticulocytosis (Table 2). One donor had a distinctly increased osmotic fragility in incubated blood and a moderate reticulocytosis (6.5%). Though not showing spherocytes in the peripheral blood smear, he probably had a mild spherocytosis which was not yet diagnosed. One donor with a " p a t h o l o g i c " AGLT showed no increased osmotic fragility when tested with hypotonic NaC1 solutions. Ninety-nine donors (6.8~ had AGLT 50 values from 5 to 30 min. These persons did not have any hematologic abnormalities, nor was the osmotic fragility, tested with hypotonic NaC1 solutions, increased. As the slightly diminished AGLT 50 value was the only abnormality, the results of AGLT 50 values between 5 and 30 min were designated as "possibly pathologic". No further examination was done with these donors.
Number of persons
(%)
16 99
(1.1) (6.8)
1349 1464
(92.1) (100)
a Using 0.0093 M sodium phosphate-buffered glycerol-saline solution, pH 6.90
A characteristic feature of erythrocytes in spherocytosis is their increased osmotic fragility. It is usually determined by the incubation of erythrocytes in hypotonic NaC1 solutions [17]. A more simple possibility for its estimation is offered by the "acidified glycerol lysis test" [23]. Using a 0.0053 M sodium phosphate buffer in the glycerolsaline solution as originally described by Zanella et al. [23], we had the same problem as that observed by Rutherford et al. [18]; i.e., there was a considerable overlap of normal controls and patients with spherocytosis. To avoid the difficulties related to the relatively low specificity of the original test we examined several different sodium phosphate-buffer concentrations for the incubation medium. Using 0.0093 M sodium phosphate buffer there was a clear cutoff between normals and patients with spherocytosis. The AGLT is critically dependent on p H and temperature control. As already described by Acharya et al. [1], we found a distinctly longer AGLT 50 at 37~ than at 25~ It is rather specific for spherocytosis, and only a few other primary or secondary erythrocyte membrane defects (severe hereditary elliptocytosis and stomatocytosis, severe autoimmunehemolytic anemia) gave a pathological result. Vettore et al. [21] proposed a modification of the AGLT using a bis-Tris buffer of distinctly higher molarity and a lower glycerol concentration than was originally published by Zanella et al. [23]. However, the
Table 2. Hematologic data and osmotic fragility of three healthy blood donors with "pathologic" AGLT Donor
Hb (g/l) Reticulocytes (%) Bilirubin (pmol/1) Osmotic fragility (in hypotonic salt solutions)
fresh blood: incub, blood:
AGLT 50 value (min)
1
2
3
156 2.0 3.4 Slightly increased
155 2.6 8.5 Slightly increased
151 6.5 6.8 Slightly increased
Slightly increased 1.5
Slightly increased 3.5
Distinctly increased 3.5
91 separation between normals and patients with spherocytosis was not as clear-cut as that achieved by the AGLT [4]. Beyond the detection of patients with spherocytosis, the AGLT was used to screen for increased osmotic fragility of the erythrocytes in healthy German blood donors. There was a wide distribution range of the osmotic fragility; 7.9% of the donors showed a continuous increase of osmotic fragility up to the range of patients with spherocytosis. The osmotic fragility might be normally distributed in the population. Our findings indicate that the donors with "pathologic" (AGLT 50 < 5 min) or "possibly pathological" results (AGLT 50: 5 - 30 rain) are not separate groups, but represent the upper end of a continuous increase of osmotic fragility. The patients with spherocytosis are characterized mostly by an AGLT 50 value of 1 - 2 min. Apparently, their increased osmotic fragility is due basically to defined structural membrane defects such as spectrin deficiency [3, 9] and ankyrin deficiency [5, 14]. The 16 healthy normals overlapping the AGLT range of spherocytosis patients had an increased hypotonic lysis of the erythrocytes and partly showed signs o f increased hemolysis. Thus, though not being a separate group, they might have a slight functional disturbance of their erythrocyte membrane resembling the defect in spherocytosis. It is unclear whether this functional defect is due to a primary defect of the erythrocyte membrane. Alternatively, it may be caused by secondary alterations such as an increased cholesterol content of the membrane or an increased sensitivity of the erythrocytes to an (extrinsic) determinant for spherocytosis, e.g., increased splenic conditioning. Some of the blood donors might be carriers of a spherocytosis trait or have a mild spherocytosis. The detection o f carriers of an erythrocyte membrane defect has major clinical implications because it may help to better explain the clinical heterogeneity of hereditary spherocytosis. By a recent evaluation of 60 families with spherocytosis a recessive manner of inheritance was proven in 20% of the families [9]. It occurred in all stages of disease severity. Previously, most families had been falsely classified as having sporadic "spherocytosis". The detection of asymptomatic carriers might help to identify families with recessive spherocytosis. Furthermore, it might well be that the distinct clinical heterogeneity of spherocytosis in some families is caused by the coinheritance of two different genetic factors, i.e., a structural membrane defect and another defect inducing an abnormal osmotic fragility o f the erythrocyte. The prevalence of these carriers could be 1.1~ in the general population. This prevalence would be 55 times higher than the supposed frequency of clinically apparent spherocytosis (0.02%) [15]. Our findings indicate that the gene frequency of spherocytosis could be as high as 0.05% in the German population.
Acknowledgements. We are indebted to Dr. H. Neumeyer, director of the blood transfusion center of the University Hospital of GOttingen, for the generous supply of blood samples from normal donors. We gratefully acknowledge the patient's consent to doing the hematological studies. Skillful computerized typewriting was done by H. Bolbeth. A portion of this work has been submitted by A. Neufeldt in the form of a dissertation in partial fulfillment of
the requirement of the Medical Faculty of the University of GOttingen. The research work was supported by the DFG grant Eb 99/1+2 and the Deutsche LeukOmie-Forschungshilfe.
References 1. Acharya J, Ferguson IL, Cassidy AG, Grimes AJ (1987) An improved acidified glycerol lysis test (AGLT) used to detect spherocytosis in pregnancy. Br J Haematol 65:343-345 2. Agre P, Casella JF, Zinkham WH, McMillan C, Bennett V (1985) Partial deficiency of erythrocyte spectrin in hereditary spherocytosis. Nature 314:380-383 3. Agre P, Asimos A, Casella JF, McMillan C (1986) Inheritance pattern and clinical response to splenectomy as a reflection of erythrocyte spectrin deficiency in hereditary spherocytes. N Engl J Med 315:1579-1583 4. Bucx MJL, Breed WPM, Hoffmann JJM (1988) Comparison of acidified glycerol lysis test, Pink test and osmotic fragility test in herditary spherocytosis: effect of incubation. Eur J Haematol 40:227-231 5. Coetzer TL, Lawler J, Liu S-C, Prchal JT, Gualtieri RJ, Brain MC, Dacie JV, Palek J (1988) Partial ankyrin and spectrin deficiency in severe, atypical hereditary spherocytosis. N Engl J Med 318:230-234 6. Dacie J (1985) The hemolytic anemias, vol. I. 3rd edn. Churchill Livingstone, Edinburgh, pp 138, 158 7. Dacie J, Lewis S (1975) Practical hematology. Churchill Livingstone, Edinburgh 8. Eber SW (1991) Correlation between the clinical severity, erythrocyte morphology and abnormal membrane skeleton in hereditary spherocytosis and elliptocytosis. Klin Paediatr 203: 284-295 9. Eber SW, Armbrust R, Schr6ter W (1990) Variable clinical severity of hereditary spherocytosis: relation to erythrocytic spectrin concentration, osmotic fragility and autohemolysis. J Pediatr 117:409-416 10. Godal H, Heisto H (1981) High prevalence of increased osmotic fragility of red blood cells in Norwegian blood donors. Scand J Haematol 27:30-34 11. Gottfried EL, Robertson NA (1974) Glycerol lysis time as a screening test for erythrocyte disorders. J Lab Clin Med 84: 746-751 12. Lux SE (1983) Disorders of the red cell membrane skeleton. In: Wyngaarden JB, Fredericksen DS, Goldstein JL, Brown MS (eds) The metabolic basis of inherited disease, 5th edn. McGraw HilI, NY, p 1573 13. Lux S, Bedrosian C, Shalev O, Morris M, Chasis J, Davies K, Savvides P, Telen M (1990) Deficiency of band 3 in dominant hereditary spherocytosis with normal spectrin content. Clin Res 38:300 (Abstract) 14. Lux SE, Tse WT, Menninger JC, John KM, Harris P, Shalev O, Chilcote RR, Marchesi SL, Watkins PC, Bennett V, McIntosh S, Collins FS, Francke U, Ward DC, Forget BG (1990) Hereditary spherocytosis associated with deletion of human erythrocyte ankyrin gene on chromosome 8. Nature 345:736-739 15. Morton NE, Mackinney AA, Kosower N, Schilling RF, Gray MP (1962) Genetics of spherocytosis. Am J Human Genet 14: 170-184 16. Nathan DG, Oski FA (1987) Hematology of infancy and childhood. Saunders, Philadelphia, p 266 17. Parpart AK, Lorenz PB, Parpart ER, Gregg JR, Chase A (1947) The osmotic resistance (fragility) of human red cells. J Clin Invest 26:636-645 18. Rutherford CJ, Postlewaight BF, Hallowes M (1986) An evaluation of the acidified glycerol lysis test. Br J Haematol 63: 119-121 19. Rybicki AC, Heath R, Wolf JL, Lubin B, Schwartz RS (1988) Deficiency of protein 4.2 in erythrocytes from a patient with a Coombs-negative hemolytic anemia. Evidence for a role of
92 protein 4.2 in stabilizing ankyrin on the membrane. J Clin Invest 81:893-901 20. SchrOter W, Kahsnitz E (1983) Diagnosis of hereditary spherocytosis in newborn infants. J Pediatr 103:460-463 21. Vettore L, Zanella A, Molaro GL, De Matteis MC, Pavesi M, Mariani M (1984) A new test for the laboratory diagnosis of spherocytosis. Acta Hematol 72:258-263
22. Winkelmann JC, Leto TL, Forget BG (1989) Spectrin genes. In: Agre P, Parker JC (eds) Red blood cell membranes. Hematology, vol 11. Decker, New York, pp 111-134 23. Zanella A, Izzo C, Rebulla P, Zanuso F, Perroni L, Sirchia G (1980) Acidified glycerol lysis test: a screening test for spherocytosis. Br J Haematol 45:481-486
Annals of
Hematology 9 Springer-Verlag 1992
Original article Prevalence of increased osmotic fragility of erythrocytes in German blood donors: screening using a modified glycerol lysis test S.W. Eber, A. Pekrun, A. Neufeldt, and W. Schr6ter Department of Pediatrics, University of G6ttingen, Robert-Koch-Strasse40, W-3400 G6ttingen, Federal Republic of Germany Received August 20, 1991/Accepted November 27, 1991
Summary. We screened for increased osmotic fragility of erythrocytes in 1464 healthy German blood donors. The osmotic fragility was determined by an acidified glycerol lysis test (AGLT) using glycerol-sodium phosphate-buffered NaC1 solution. Since the original test described by Zanella et al. [23] showed only low specificity for hereditary spherocytosis, we used a modification with 0.0093 M sodium phosphate-buffered glycerol-saline solution, pH 6.90, instead of the original 0.0053 M sodium phosphate buffer, pH 6.85. Sixteen of the donors (1.1%) had a "pathologic result," similar to that of 32 patients with hereditary spherocytosis: AGLT 50 < 5 rain ("half-time of AGLT, defining normal and pathologic results). The osmotic fragility of the erythrocytes from 12 of these donors was further investigated using the conventional test with hypotonic NaC1 solutions. With one exception, increased osmotic fragility was verified in all of them by both tests. Further hematologic data showed a mild reticulocytosis (2% and 2.6%) in two of the donors. One donor had a moderate reticulocytosis of 6.5%, probably due to a mild, previously undiagnosed spherocytosis; 99 of the donors had an intermediate result (AGLT 50: 5 - 30 min). Hypotonic lysis of their erythrocytes by the conventional method showed a normal result; there were no signs of increased hemolysis. Thus they are not definitely regarded as having increased osmotic fragility of their erythrocytes. Erythrocyte osmotic fragility shows a wide distribution range in the normal population and might be normally distributed. Thus the blood donors with "pathologic AGLT ( < 5 min)" probably represent only one end of a continuum of salt-dependent hemolysis, and not a separate entity. However, they did show additional minor signs of a functional defect of the erythrocyte membrane and therefore could be carriers of a spherocytosis trait. The frequency of carriers of an erythrocyte membrane defect (possible spherocytosis trait) could be as high as 1.1% in the general population and would distinctly exceed the prevalence of patients with apparent spherocytosis (0.02%). Address for correspondence: S.W. Eber
Key words: Osmotic fragility - Acidified glycerol lysis test - Hemolytic anemia - Hereditary spherocytosis
Introduction Hereditary spherocytosis is the most common congenital hemolytic anemia in middle Europe. Its incidence in the Caucasian population was estimated to be 200-300 per million [6, 15]. These figures are probably an underestimate, because about a quarter of the patients show a mild clinical course [12] and therefore the diagnosis of spherocytosis may be missed. Hemolytic anemia ranges from mild, fully compensated forms to severe recessive forms, requiring regular transfusions [9]. In 20% of the patients the manner of inheritance is recessive, with both parents being asymptomatic, heterozygous carriers [2, 3, 9]. A high prevalence of carriers with a symptomless trait for spherocytosis was reported by Dacie [6] and by Godal and Heisto [10]. Spherocytosis seems to be a heterogeneous clinical entity, and primary defects of spectrin [22], ankyrin [14], band 3 [13], and band 4.2 [19] have been described. The determination of the osmotic fragility of the erythrocytes is widely used as a diagnostic tool for spherocytosis. Therefore, it seemed appropriate to use a rapid and sensitive test of osmotic fragility in a population screening for spherocytosis trait. However, measuring the osmotic fragility in hypotonic NaC1 solution [17] is time consuming, and diagnosis fails in many of the mild cases [6]. Gottfried and Robertson [11] used the lysis time of erythrocytes in isotonic glycerol solution, pH 7.4. Because of the relatively low sensitivity of this "glycerol lysis test" Zanella et al. [23] proposed the "acidified glycerol lysis test" (AGLT) using 0.0053 M sodium phosphate-buffered glycerol-saline solution, pH 6.85. They claimed 100% sensitivity for spherocytosis. Bucx et al. [4] found the same sensitivity using 24-h incubated blood instead of fresh blood. However, in a re-evaluation, Rutherford et al. [18] found a considerable overlap between
89 spherocytosis p a t i e n t s a n d controls. We therefore m o d i fied the A G L T by using a higher s o d i u m p h o s p h a t e - b u f fer c o n c e n t r a t i o n a n d a slightly increased p H value. W i t h this m o d i f i c a t i o n b o t h high specificity a n d high sensitivity were achieved, w h i c h allowed for p o p u l a t i o n screening. We tested 1464 n o r m a l b l o o d d o n o r s . T h e o s m o t i c fragility o f erythrocytes f r o m d o n o r s with " p a t h o l o g i c " A G L T results was f u r t h e r investigated with the conventional test, using h y p o t o n i c salt s o l u t i o n s [17]. A n increased o s m o t i c h e m o l y s i s was f o u n d for m o s t o f t h e m . A d d i tionally, the b l o o d d o n o r s were tested for m i n o r signs o f increased hemolysis. It was thus p o s s i b l e to evaluate the f r e q u e n c y o f clinically i n a p p a r e n t carriers for spherocytosis.
Methods
Patients and normal blood donors A total of 1464 blood donors (1127 male) to the blood transfusion center of the University Hospital of G0ttingen were investigated. The donors were healthy and had no signs of infection. Complete blood cell count, liver enzymes, and renal function were normal. Serologic tests for viral hepatitis and lues were negative. Most donors were 20-31 years old (medium age 25 years). In addition, 44 German patients (19 male) with spherocytosis at the Department of Pediatrics, University of GOttingen, and some other hospitals were investigated. The age of the children was 6 months to 18 years. All patients had hemolytic anemia of varying degree; diagnosis was made by the detection of spherocytes in the blood smear and proof of increased osmotic fragility of the erythrocytes. A reduced concentration of spectrin (below 80% of normal) in the erythrocyte membrane was found in nearly all patients with moderate to severe hemolytic anemia (hemoglobin concentration < 120 g/l, reticulocyte count > 6%, bilirubin concentration > 34 amol/1). Patients with mild disease had a compensated hemolytic anemia (hemoglobin concentration 110-150 g/l, reticulocyte count < 6%, bilirubin concentration up to 34 amol/1) and a normal spectrin concentration [9].
5 ml phosphate-buffered saline (PBS; 0.01-0.05 M sodium phosphate buffer, pH 6.90, containing 0.135 MNaC1). The pH value of the sodium phosphate buffer was adjusted to pH 6.90 _+ 0.05 at 25~ by mixing primary and secondary sodium phosphate solutions containing the desired molarity without further addition of NaOH or HC1 in order to avoid a rise of the osmolarity. To 1 ml of this suspension 2 ml of glycerol solution (glycerol 0.3 M, sodium chloride 0.041 M, sodium phosphate buffer 0.003-0.015 M, pH 6.90, corresponding to the different sodium phosphate concentrations in PBS) was added. The final concentrations in the assay were as follows: 0.0053-0.0265 M sodium phosphate, 0.072 M NaC1, 0.2 M glycerol; final osmolarities ranged from 0.356 to 0.406 osmol/1. After the initial evaluation of specificity and sensitivity of the test, all further measurements were done with 0.0093 M sodium phosphate-buffered glycerol-saline solution, pH 6.90 (final osmolarity: 0.366 osmol/1). The rate of hemolysis was followed by measuring the decline of "extinction" at 625 nm, which is due mainly to the turbidity of the suspended erythrocytes. Half-time of the test (AGLT 50) was defined as the time required for the "extinction" to fall to half of the initial value. To avoid sedimentation of the erythrocytes during the 30-rain incubation, the cuvettes were shaken gently every 10 rain.
Results
Sensitivity and specificity of the acidified glycerol lysis test T h e o r i g i n a l test d e s c r i b e d by Z a n e l l a et al. [23] using 0.0053 M p h o s p h a t e - b u f f e r e d glycerol-saline s o l u t i o n gave a c o n s i d e r a b l e overlap b e t w e e n p a t i e n t s with spherocytosis a n d n o r m a l controls (Fig. 1). Five o f the n o r m a l c o n t r o l values fell below 5 rain. Therefore, we c h a n g e d the c o n c e n t r a t i o n o f s o d i u m p h o s p h a t e b u f f e r until b o t h m a x i m a l sensitivity a n d specificity were r e a c h e d (Fig. 2). Using a s o d i u m p h o s p h a t e - b u f f e r c o n c e n t r a t i o n o f 0.0093 M , 32 p a t i e n t s with spherocytosis consistently h a d an A G L T 50 value below 5 m i n (15 s - 4 . 5 min, with a m a x i m u m frequency between 1 a n d 2 min, d a t a n o t shown).
Hematologic data, osmotic fragility Hematologic data were obtained by routine methods according to Dacie and Lewis [7]. Normal hemoglobin concentrations of agematched children were taken [16]; normal reticulocyte range was 0.5%-1.5%, normal total bilirubin concentration less than 17 pmol/1. Spectrin concentration of the erythrocytes was determined by means of an ELISA using monoclonal spectrin antibodies (normal mean _+SD: 226 _+ 35 xl03 copies per cell) [8]. Osmotic fragility of erythrocytes in hypotonic NaC1 solutions was determined in fresh and 24-h-incubated blood using the method of Parpart et al. [171. For a semiquantitative evaluation, the NaC1 concentration (g/dl) was determined at the onset of osmotic hemolysis (1070hemolysis), at the median (50~ hemolysis), and at completion (100~ hemolysis). Normal ranges (two standard deviations) were as previously published [20]. The osmotic fragility in fresh or incubated blood samples of the patients was designated as slightly increased if one or two of the measured values were outside the normal range; if all three values were outside the normal limits, the osmotic fragility was described as distinctly increased. AGLT was done according to the method of Zanella et al. [23] at 25~ hut using different sodium phosphate-buffered glycerolsaline solutions. The pH value was increased to pH 6.9 +_ 0.05. Control of pH as well as of temperature is rather critical: decrease of the temperature to 20~ gave a shortened half-time of the test (minimum 15 min), increase to 37~ a distinctly prolonged halftime. EDTA-anticoagulated whole blood (20 ~1) was mixed with
Screening for increased osmotic fragility of healthy blood donors We tested 1464 h e a l t h y b l o o d d o n o r s . T h e results are given in Table 1. T h e m a j o r i t y o f the d o n o r s h a d n o r m a l A G L T 50 values above 30 min. Sixteen d o n o r s h a d a clear decrease o f A G L T 50 below 5 min, o v e r l a p p i n g with t h a t o f p a t i e n t s with spherocytosis. T h e result o f the A G L T is therefore d e s i g n a t e d as " p a t h o l o g i c " . Twelve o f the 16 d o n o r s were f u r t h e r evaluated. Clinical h i s t o r y was uneventful, giving no hint for a h e m o l y t i c a n e m i a in the f a m i l y history; the spleen was n o t increased. T h e o s m o t i c fragility o f their erythrocytes was tested with h y p o t o n i c NaC1 s o l u t i o n s in fresh a n d i n c u b a t e d b l o o d samples. E i g h t o f these d o n o r s h a d slightly increased o s m o t i c fragility o n l y after 24-h i n c u b a t i o n o f b l o o d samples. T h e i r h e m o g l o b i n c o n c e n t r a t i o n (range 1 3 4 - 1 6 3 g / l ) a n d bilir u b i n c o n c e n t r a t i o n (range 3 . 4 - 1 0 . 2 p m o l / 1 ) were n o r mal; the reticulocyte counts were n o t increased (range 0 . 3 % - 1 % ) . Two o f the d o n o r s with " p a t h o l o g i c " A G L T h a d a slightly increased o s m o t i c fragility b o t h in fresh a n d in i n c u b a t e d b l o o d samples. B o t h h a d n o r m a l h e m o -
90
Normal controls
9oeee 9149
Hereditary spherocytosis
ii
9
eee
9 9149 9
9 1 4 9 1 4 9 1 49 9 1e4e99 1 4 9 1 4 9
~9 9
L 0
I 5
I I I 10 15 20 AGLT 5o (minutes)
I 25
.t/" 30
Fig. 1. AGLT results of 14 patients with spherocytosis and 186 normal children (age 1-18 years) using 0.0053 M phosphate-buffered glycerol-saline solution, pH 6.85. ;~'s~%indicates 155 AGLT 50 values above 30 min "" 9
Specificity (%)
Sensitivity (%)
100
100
I
l
50
0
PN 2,7
Discussion 50
0 PN PN PN PN PN PN PN PN 5~3 6~6 8~0 9~3 10~6 13,3 15,9 26,5 [ram phosph{atebuffer )
Fig. 2. Sensitivity and specificity of the AGLT at various sodium phosphate-buffer concentrations in the glycerol-saline solution (final concentrations): P, Patients (n = 9); N, normal controls (n = 18). Sensitivity (hatched bars) is indicated by the percentage of patients with spherocytosis giving a pathologic result (AGLT 50 value < 5 rain), specificity (blank bars) by the percentage of normals showing a normal result (AGLT 50 value > 30 min)
Table 1. Acidified glycerol lysis test in healthy blood donors a AGLT 50 (min)
Classification
0- 5 5 - 30
Pathologic Possibly pathologic Normal
> 30 Total
globin concentrations but disclosed a mild reticulocytosis (Table 2). One donor had a distinctly increased osmotic fragility in incubated blood and a moderate reticulocytosis (6.5%). Though not showing spherocytes in the peripheral blood smear, he probably had a mild spherocytosis which was not yet diagnosed. One donor with a " p a t h o l o g i c " AGLT showed no increased osmotic fragility when tested with hypotonic NaC1 solutions. Ninety-nine donors (6.8~ had AGLT 50 values from 5 to 30 min. These persons did not have any hematologic abnormalities, nor was the osmotic fragility, tested with hypotonic NaC1 solutions, increased. As the slightly diminished AGLT 50 value was the only abnormality, the results of AGLT 50 values between 5 and 30 min were designated as "possibly pathologic". No further examination was done with these donors.
Number of persons
(%)
16 99
(1.1) (6.8)
1349 1464
(92.1) (100)
a Using 0.0093 M sodium phosphate-buffered glycerol-saline solution, pH 6.90
A characteristic feature of erythrocytes in spherocytosis is their increased osmotic fragility. It is usually determined by the incubation of erythrocytes in hypotonic NaC1 solutions [17]. A more simple possibility for its estimation is offered by the "acidified glycerol lysis test" [23]. Using a 0.0053 M sodium phosphate buffer in the glycerolsaline solution as originally described by Zanella et al. [23], we had the same problem as that observed by Rutherford et al. [18]; i.e., there was a considerable overlap of normal controls and patients with spherocytosis. To avoid the difficulties related to the relatively low specificity of the original test we examined several different sodium phosphate-buffer concentrations for the incubation medium. Using 0.0093 M sodium phosphate buffer there was a clear cutoff between normals and patients with spherocytosis. The AGLT is critically dependent on p H and temperature control. As already described by Acharya et al. [1], we found a distinctly longer AGLT 50 at 37~ than at 25~ It is rather specific for spherocytosis, and only a few other primary or secondary erythrocyte membrane defects (severe hereditary elliptocytosis and stomatocytosis, severe autoimmunehemolytic anemia) gave a pathological result. Vettore et al. [21] proposed a modification of the AGLT using a bis-Tris buffer of distinctly higher molarity and a lower glycerol concentration than was originally published by Zanella et al. [23]. However, the
Table 2. Hematologic data and osmotic fragility of three healthy blood donors with "pathologic" AGLT Donor
Hb (g/l) Reticulocytes (%) Bilirubin (pmol/1) Osmotic fragility (in hypotonic salt solutions)
fresh blood: incub, blood:
AGLT 50 value (min)
1
2
3
156 2.0 3.4 Slightly increased
155 2.6 8.5 Slightly increased
151 6.5 6.8 Slightly increased
Slightly increased 1.5
Slightly increased 3.5
Distinctly increased 3.5
91 separation between normals and patients with spherocytosis was not as clear-cut as that achieved by the AGLT [4]. Beyond the detection of patients with spherocytosis, the AGLT was used to screen for increased osmotic fragility of the erythrocytes in healthy German blood donors. There was a wide distribution range of the osmotic fragility; 7.9% of the donors showed a continuous increase of osmotic fragility up to the range of patients with spherocytosis. The osmotic fragility might be normally distributed in the population. Our findings indicate that the donors with "pathologic" (AGLT 50 < 5 min) or "possibly pathological" results (AGLT 50: 5 - 30 rain) are not separate groups, but represent the upper end of a continuous increase of osmotic fragility. The patients with spherocytosis are characterized mostly by an AGLT 50 value of 1 - 2 min. Apparently, their increased osmotic fragility is due basically to defined structural membrane defects such as spectrin deficiency [3, 9] and ankyrin deficiency [5, 14]. The 16 healthy normals overlapping the AGLT range of spherocytosis patients had an increased hypotonic lysis of the erythrocytes and partly showed signs o f increased hemolysis. Thus, though not being a separate group, they might have a slight functional disturbance of their erythrocyte membrane resembling the defect in spherocytosis. It is unclear whether this functional defect is due to a primary defect of the erythrocyte membrane. Alternatively, it may be caused by secondary alterations such as an increased cholesterol content of the membrane or an increased sensitivity of the erythrocytes to an (extrinsic) determinant for spherocytosis, e.g., increased splenic conditioning. Some of the blood donors might be carriers of a spherocytosis trait or have a mild spherocytosis. The detection o f carriers of an erythrocyte membrane defect has major clinical implications because it may help to better explain the clinical heterogeneity of hereditary spherocytosis. By a recent evaluation of 60 families with spherocytosis a recessive manner of inheritance was proven in 20% of the families [9]. It occurred in all stages of disease severity. Previously, most families had been falsely classified as having sporadic "spherocytosis". The detection of asymptomatic carriers might help to identify families with recessive spherocytosis. Furthermore, it might well be that the distinct clinical heterogeneity of spherocytosis in some families is caused by the coinheritance of two different genetic factors, i.e., a structural membrane defect and another defect inducing an abnormal osmotic fragility o f the erythrocyte. The prevalence of these carriers could be 1.1~ in the general population. This prevalence would be 55 times higher than the supposed frequency of clinically apparent spherocytosis (0.02%) [15]. Our findings indicate that the gene frequency of spherocytosis could be as high as 0.05% in the German population.
Acknowledgements. We are indebted to Dr. H. Neumeyer, director of the blood transfusion center of the University Hospital of GOttingen, for the generous supply of blood samples from normal donors. We gratefully acknowledge the patient's consent to doing the hematological studies. Skillful computerized typewriting was done by H. Bolbeth. A portion of this work has been submitted by A. Neufeldt in the form of a dissertation in partial fulfillment of
the requirement of the Medical Faculty of the University of GOttingen. The research work was supported by the DFG grant Eb 99/1+2 and the Deutsche LeukOmie-Forschungshilfe.
References 1. Acharya J, Ferguson IL, Cassidy AG, Grimes AJ (1987) An improved acidified glycerol lysis test (AGLT) used to detect spherocytosis in pregnancy. Br J Haematol 65:343-345 2. Agre P, Casella JF, Zinkham WH, McMillan C, Bennett V (1985) Partial deficiency of erythrocyte spectrin in hereditary spherocytosis. Nature 314:380-383 3. Agre P, Asimos A, Casella JF, McMillan C (1986) Inheritance pattern and clinical response to splenectomy as a reflection of erythrocyte spectrin deficiency in hereditary spherocytes. N Engl J Med 315:1579-1583 4. Bucx MJL, Breed WPM, Hoffmann JJM (1988) Comparison of acidified glycerol lysis test, Pink test and osmotic fragility test in herditary spherocytosis: effect of incubation. Eur J Haematol 40:227-231 5. Coetzer TL, Lawler J, Liu S-C, Prchal JT, Gualtieri RJ, Brain MC, Dacie JV, Palek J (1988) Partial ankyrin and spectrin deficiency in severe, atypical hereditary spherocytosis. N Engl J Med 318:230-234 6. Dacie J (1985) The hemolytic anemias, vol. I. 3rd edn. Churchill Livingstone, Edinburgh, pp 138, 158 7. Dacie J, Lewis S (1975) Practical hematology. Churchill Livingstone, Edinburgh 8. Eber SW (1991) Correlation between the clinical severity, erythrocyte morphology and abnormal membrane skeleton in hereditary spherocytosis and elliptocytosis. Klin Paediatr 203: 284-295 9. Eber SW, Armbrust R, Schr6ter W (1990) Variable clinical severity of hereditary spherocytosis: relation to erythrocytic spectrin concentration, osmotic fragility and autohemolysis. J Pediatr 117:409-416 10. Godal H, Heisto H (1981) High prevalence of increased osmotic fragility of red blood cells in Norwegian blood donors. Scand J Haematol 27:30-34 11. Gottfried EL, Robertson NA (1974) Glycerol lysis time as a screening test for erythrocyte disorders. J Lab Clin Med 84: 746-751 12. Lux SE (1983) Disorders of the red cell membrane skeleton. In: Wyngaarden JB, Fredericksen DS, Goldstein JL, Brown MS (eds) The metabolic basis of inherited disease, 5th edn. McGraw HilI, NY, p 1573 13. Lux S, Bedrosian C, Shalev O, Morris M, Chasis J, Davies K, Savvides P, Telen M (1990) Deficiency of band 3 in dominant hereditary spherocytosis with normal spectrin content. Clin Res 38:300 (Abstract) 14. Lux SE, Tse WT, Menninger JC, John KM, Harris P, Shalev O, Chilcote RR, Marchesi SL, Watkins PC, Bennett V, McIntosh S, Collins FS, Francke U, Ward DC, Forget BG (1990) Hereditary spherocytosis associated with deletion of human erythrocyte ankyrin gene on chromosome 8. Nature 345:736-739 15. Morton NE, Mackinney AA, Kosower N, Schilling RF, Gray MP (1962) Genetics of spherocytosis. Am J Human Genet 14: 170-184 16. Nathan DG, Oski FA (1987) Hematology of infancy and childhood. Saunders, Philadelphia, p 266 17. Parpart AK, Lorenz PB, Parpart ER, Gregg JR, Chase A (1947) The osmotic resistance (fragility) of human red cells. J Clin Invest 26:636-645 18. Rutherford CJ, Postlewaight BF, Hallowes M (1986) An evaluation of the acidified glycerol lysis test. Br J Haematol 63: 119-121 19. Rybicki AC, Heath R, Wolf JL, Lubin B, Schwartz RS (1988) Deficiency of protein 4.2 in erythrocytes from a patient with a Coombs-negative hemolytic anemia. Evidence for a role of
92 protein 4.2 in stabilizing ankyrin on the membrane. J Clin Invest 81:893-901 20. SchrOter W, Kahsnitz E (1983) Diagnosis of hereditary spherocytosis in newborn infants. J Pediatr 103:460-463 21. Vettore L, Zanella A, Molaro GL, De Matteis MC, Pavesi M, Mariani M (1984) A new test for the laboratory diagnosis of spherocytosis. Acta Hematol 72:258-263
22. Winkelmann JC, Leto TL, Forget BG (1989) Spectrin genes. In: Agre P, Parker JC (eds) Red blood cell membranes. Hematology, vol 11. Decker, New York, pp 111-134 23. Zanella A, Izzo C, Rebulla P, Zanuso F, Perroni L, Sirchia G (1980) Acidified glycerol lysis test: a screening test for spherocytosis. Br J Haematol 45:481-486
