Variable clinical severity of hereditary spherocytosis: Relation to erythrocytic spectrin concentration, osmotic fragility, and autohemolysis Stefan W. Eber, MD, Ralph Armbrust, MD, a n d Werner SchrOter, MD From the Department of Pediatrics, University Hospital of G~ttingen, Federal Republic of Germany To determine whether stratifying hereditary spherocytosis by degree of severity could provide guidelines regarding which patients would benefit from splenectomy, we evaluated the clinical characteristics of 80 patients (63 children) and 27 healthy relatives. In addition to routine hematologic determinations, osmotic fragility, autohemolysis, erythrocyte spectrin content, and erythrocyte membrane lipid phoshorus were measured and correlated with the disease severity. Four categories were identified: (I) spherocytosis as a trait in symptom-free relatives of patients with recessively inherited disease; (2) mild and (3) moderate spherocytosis, largely observed in patients with dominantly inherited disease; and (4) severe spherocytosis, observed in only two patients, who were characterized by recessive inheritance and transfusion dependence. By the identification of carriers, a recessive mode of inheritance could be demonstrated in 20% of the families with spherocytosis. The eryfhrocyte spectrin concentration was normal in carriers and patients with mild spherocytosis, and was significantly reduced in the moderate and severe states of the disease. This difference was not accounted for by reduced membrane area of the cells, as measured by the phospholipid concentration per cell. We conclude that patients with mild spherocytosis usually do not require splenectomy durlng childhood and adolescence; patients with moderate or severe disease should have splenectomy. Patients with severe spherocytosis have a partial response to splenectomy but a considerable degree of increased hemolysis persists. Most patients with less than 80% of normal spectrin content require splenectomy. (J PEDIATR1990;117:409-
16)
Among the erythocyte membrane defects, hereditary spherocytosis is the most common. 1 The primary biochemical defect, although not definitely identified in most cases, is surmised to reside in the erythrocyte membrane skeleton, a proteinaceous network of spectrin, actin, band 4.1, and Supported by grant No. Eb 99/1-1 from the Deutsche Forschungsgemeinschaft. Submitted for publication Aug. 22, 1989; accepted April 11, 1990. Reprint requests: Priv. Doz. Dr. S. W. Eber, Department of Pediatrics, University Hospital of G6ttingen, Robert-Koch Strasse 40, D-3400 G6ttingen, Federal Republic of Germany. 9/20/21526
minor proteins that covers the inner surface of the plasma membrane. 2-7 Agre et al. 3, 4 described a deficiency of spectrin in patients with dominantly inherited spherocytosis and in those with the recessively inherited form. However, differential diagnosis may be difficult, especially for patients with mild spherocytosis, who often lack some of the cardinal features, and the detection of clinically unaffected heterozygous carriers in families with a supposed recessive form has generally been impossible by conventional diagnostic criteria. The clinical symptoms are variable, and classification according to clinical severity has rarely been attempted. Quantitation of hematologic characteristics, osmotic fra409
4 10
Eber, Armbrust, and Schrg)ter
The Journal of Pediatrics September 1990
Table I. Hematologic and clinical data from 107 patients with spherocytosis
Classification
Hemoglobin (gm/L)'~ Reticulocytes (%)'~ Bilirubin 0zmol/L)t Reticulocyte production index~,:~ Postnatal exchange transfusionw Transfusion-dependent crisesw Splenectomyw
Carrier (n ~ 26)
149 • 14 2.1 --- 0.8 9.0 • 3.7 1.33 • 0.47 0 0 0
Mild spherocytosis (n = 30)
128 4.7 27.2 2.44
• 15 +_ 1.9 _+ 14.1 • 0.63 2
911 3
Moderate spherocytosis (n = 49)
108 14.2 41.1 6.28
_+ 17 • 4.7 • 16.0 • 2.26 10 20 22
Severe spherocytosis ~ (n = 2)
60-80 l 1.4 63.0 ND 1 RT 1
ND, Not determinedbecause of regular transfusions;RT, regular transfusions. *Data before transfusion. tValues are expressed as mean -+ SD. :~Normalmean +- SD: 1 • 0.25, w of patients. IIOnlyone crisis per patient.
Table II. Inheritance pattern in four different classes of spherocytosis Dominant
Trait* (n = 26) Mild spherocytosis (n = 30) Moderate spherocytosis (n = 49) Severe spherocytosis (n = 2)
Recessive
Spontaneous
Not k n o w n
0
24
0
2
22
5
1
2
36
7
4
2
0
2
0
0
*Note that carriers with the trait are always related to patients with recessivelyinheritedspherocytosis gility, autohemolysis, and spectrin deficiency in relation to the severity of spherocytosis has been done in a few studies, none of them comprehensive. 3, 8-10 In addition, recommendations about which groups of patients should have splenectomy differ widely. Some authors argue for splenectomy in nearly all cases, 11-t3 but because of the prevalence of postsplenectomy infections, 14, t5 such a broad indication cannot be accepted. Quantitative laboratory criteria indicating the need for splenectomy have not been laid down. We have attempted to develop such criteria on the basis of clinical history, hematologic data, autohemolysis, and osmotic fragility for a sample of patients with spherocytosis, some of whom had spectrin concentration determinations.
METHODS Eighty patients (45 female patients) with the clinical history of spherocytosis were evaluated; 63 patients were children ranging from 4 weeks of age to 15 years, and 17 were adults. All patients were referred to our hospital from 1974 to 1988 because of hemolytic anemia. The diagnosis of spherocytosis was based on the increased osmotic fragility and increased autohemolysis of the erythrocytes, as well
as on the appearance of spherocytes in the peripheral blood smear. The spleen was enlarged to a varying degree. The result of a direct Coombs test was negative in all cases. In addition to the 80 patients, we investigated 27 healthy relatives who had minor signs of spherocytosis such as slightly increased reticulocyte count, osmotic fragility, and autohemolysis. The control sample consisted of healthy children and adults with no signs of increased hemolysis; their ages ranged from 5 to 30 years. Hematology. Hematologic data were determined with a Microcell Counter Sysmex CC-180 (Toa Medical Electronics, Kobe, Japan). The reticulocyte production index was calculated according to Koepke and Koepke.16 The osmotic fragility of the erythrocytes was measured in fresh blood and after 24 hours of incubation, as described by Parpart et al. 17 For a semiquantitative evaluation, the sodium chloride concentrations (gm/dl) at the onset of osmotic hemolysis (1% hemolysis), at the median (50% hemolysis), and at completion (100% hemolysis) were determined. Normal ranges ( + 2 SD) were as previously published. 18 The osmotic fragility in fresh or incubated blood 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
Volume 117 Number 3
Severity of hereditary spherocytosis
411
Fig. t. E•icacy•fsplenect•myinpatientswithmi•d•m•derate•andseverespher•cyt•sis•Hem•gl•binc•ncentrati•n (A) and reticulocyte counts (B) are given before (hatched bars) and after (stippled bars) splenectomy. Mean value and 1 SD range are indicated. HS, Hereditary spherocytosis.
Table III. Spectrin and lipid phosphorus in spherocytosis Spectrin molecules (• 103/cell) Normal (n = 13) Trait (n = 4) Mild spherocytosis (n = 7) Moderate spherocytosis* (n = 6) Severe spherocytosis* (n = 2)
226 _+ 54 244 _+ 77 229 _+ 65"-[ 162 _+ 46 (72%) 142 (62%)
A
Lipid phosphorus (10 -16 mol/cell)
Spectrin per lipid phosphorus (mmol/mol)
3.3 _+ 0.4 3.0 _+ 0.3 3.1 _+ 0.6
1.14 _+ 0.29 1.28 _+ 0.26 1.26 _+ 0.37
3.0 +_ 0.4
0.89 _+ 0.20 (78%) 0.8 (70%)
3.0
Values (except for severe spherocytosis) are expressed as mean +_ SD. Bracket indicates signficance level of c~ = 0.1. *Percentage of the normal value is given in parentheses.
osmotic fragility was described as distinctly increased. Autohemolysis was measured without and with the addition of glucose according to Young et al. 19 The reduction of autohemolysis (percentage of correctability) by the addition of glucose was calculated according to the following formula: t Correctability (%) = 100 X
1-
Autohemolysis with glucose (%) ] Autohemolysis ] without glucose (%),/
Statistical analysis. Statistical evaluations were performed with the Wilcoxon-Mann-Whitney test. 2~ Concentration of speetrin and lipid phosphorus in erythrocytes. Erythrocytes were washed in isotonic sodium chloride solution and pressed through cotton wool to remove leukocytes. For the preparation of lysed erythrocytes, up to 0.5 ml
packed cells (5 X 109 cells) were lysed in the same volume of 0.05 m o l / L tris-(hydroxymethyl)-aminomethane-HC1 buffer, pH 8.2, containing 0.38 m o l / L glycerol, 10% wt/vol sodium dodecyl-sulfate, and 10% vol/vol mercaptoethanol. The samples were incubated at 30 ~ C for 15 minutes, diluted with double distilled water (1 "5), and heated at 100 ~ C for 7 minutes. After centrifugation (10 minutes at 48,000 g) the supernatant was diluted 1:2 with sample buffer. Polyacrylamide slab gel electrophoresis in the presence of sodium dodecyl sulfate of lysed erythrocytes, together with spectrin standards of known concentration, was performed according to Fairbanks et al.21; gels were stained with Coomassie brilliant blue G 250. After being destained, the spectrin bands were cut out. The dye was extracted in 25% pyridine and qiaantitated spectrophotometrically. 22 Spectrin concentration was calculated as copies of a~3 dimer per lysed erythrocyte. For measurement of lipid phosPhorus in
4 12
Eber, Armbrust, and Schrg)ter
The Journal of Pediatrics September 1990
Table IV. Osmotic fragility in spherocytosis Fresh blood (gm NaCl/dl) Osmotic hemolysis (%)
1
50
Incubated blood at 24 hr (gm NaCl/dl) 100
I
50
Normal (n = 13) 0.53 + 0.02 0.43 _+ 0.02 0.17 _+ 0.03]]-] 0.75 _+ 0.05 0.46 _+ 0.03 Trait (n = 26) 0.53 _+ 0.04---I 0.45 _+ 0.04 0.33 + 0.08-a[[ 0.73 _+ 0.09--] 0.55 _+ 0.03 Mild spherocytosis 0.57 + 0'06~------a/ 0.46 _+ 0.03 0.36 + 0.06-a[ 0.88 _+ 0.06--1 0.62 _+ 0.06 (n = 20) / 1 Moderate spherocytosis 0.69 + 0.09---1 0.49 _+ 0.04 0.38 _+ 0.07 ~ - 0.98 _+ 0.08 0.67 _+ 0.08 (n = 30) Severe spherocytosis 0.67 0.49 0.40 0.92 0.59 (n = 2) Values (except for severe spherocytosis)are expressed as mean _+standard deviation.Brackets indicatesignificantdifferences (a = 0.05).
erythrocytes, cholesterol and phospholipids from up to 1 ml packed cells (10 l~ cells) were extracted according to the method of Rose and Oklander. 23 Lipid extracts were ashed at 300 ~ C, and the lipid phosphorus concentration per erythrocyte was determined by the ammonium molybdate color reaction. 24 Spectrin density was defined as the concentration of moles of spectrin per mole of lipid phosphorus. RESULTS
Classification of the patients according to clinical severity. The patients could be separated into three classes of differing clinical severity (mild, moderate, and severe); a fourth class consisted of symptom-free carriers. The criteria for the classification were (1) the concentration of hemoglobin; (2) signs of increased hemolysis (reticulocyte counts, reticulocyte production index, serum concentration of bilirubin); (3) the frequency of severe neonatal hyperbilirubinemia requiring exchange transfusion; and (4) the number of transfusion-dependent aplastic or hemolytic crises (Table I). Heterozygous carriers. Heterozygous carriers were detected exclusively by family screening of patients with symptomatic sphcrocytosis. None of the carriers had any clinical symptoms. Only by considering all available data (rcticulocyte count, concentration of haptoglobin, osmotic fragility, and autohemolysis) could the presence of a genetic abnormality be demonstrated; sphcrocytes were not found. Mild spherocytosis. Patients with mild spherocytosis had a compensated hemolytic anemia. Rcticulocyte counts wcrc 6%, and erythropoiesis was stimulated more than threefold. The serum bilirubin concentration usually exceeded 34 m m o l / L (2 mg/dl), and 93% of the patients had spherocytes. A postnatal exchange transfusion was necessary in 20% of the cases. About half of the patients had had several hemolytic crises and one aplastic crisis, and had required several transfusions. Four patients had gallstones, which were removed by choleeystectomy at the time of the splenectomy. Splenectomy was performed on about half of the patients. Severe spherocytosis. Two unrelated patients had a severe form requiring regular transfusions from birth to sustain a hemoglobin level greater than 60 gm/L. The parents of both patients were carriers of spherocytosis (mild reticulocytosis, slightly elevated osmotic fragility, and raised autohemolysis). Thus these two patients had recessively inherited spherocytosis. One of them already had gallstones at the age of 5 years. Inheritance pattern. The inheritance patterns are given in Table II. The patients belonged to 60 unrelated families. Forty families (66%) had a dominant mode and 12 families (20%) a recessive mode of inheritance. In the latter families, both parents were symptom-free carriers. In four families a spontaneous mutation was inferred, and in another four families no genealogic studies could be performed. Response to splenectomy. Splenectomy was effective, leading to a normal hemoglobin concentration in all patients (Fig. 1). Many patients with moderate spheroeytosis had mild reticulocytosis, indicative of a continuing slightly increased hemolysis after splenectomy. With rare exceptions the serum bilirubin concentration became normal after splenectomy in all patients with mild and moderate spherocytosis. The patient with severe spherocytosis had mild reticulocytosis and a slightly increased bilirubin concentration (19 ~ m o l / L = 1.1 mg/dl) after splenectomy. On average, the three patients with mild spherocytosis had
Volume 117 Number 3
Severity o f hereditary spherocytosis
4 13
,= 33
A
B n=22
28-
9024 ~.
n =22
~ 0
20-
E O r o
16-
16)
Among the erythocyte membrane defects, hereditary spherocytosis is the most common. 1 The primary biochemical defect, although not definitely identified in most cases, is surmised to reside in the erythrocyte membrane skeleton, a proteinaceous network of spectrin, actin, band 4.1, and Supported by grant No. Eb 99/1-1 from the Deutsche Forschungsgemeinschaft. Submitted for publication Aug. 22, 1989; accepted April 11, 1990. Reprint requests: Priv. Doz. Dr. S. W. Eber, Department of Pediatrics, University Hospital of G6ttingen, Robert-Koch Strasse 40, D-3400 G6ttingen, Federal Republic of Germany. 9/20/21526
minor proteins that covers the inner surface of the plasma membrane. 2-7 Agre et al. 3, 4 described a deficiency of spectrin in patients with dominantly inherited spherocytosis and in those with the recessively inherited form. However, differential diagnosis may be difficult, especially for patients with mild spherocytosis, who often lack some of the cardinal features, and the detection of clinically unaffected heterozygous carriers in families with a supposed recessive form has generally been impossible by conventional diagnostic criteria. The clinical symptoms are variable, and classification according to clinical severity has rarely been attempted. Quantitation of hematologic characteristics, osmotic fra409
4 10
Eber, Armbrust, and Schrg)ter
The Journal of Pediatrics September 1990
Table I. Hematologic and clinical data from 107 patients with spherocytosis
Classification
Hemoglobin (gm/L)'~ Reticulocytes (%)'~ Bilirubin 0zmol/L)t Reticulocyte production index~,:~ Postnatal exchange transfusionw Transfusion-dependent crisesw Splenectomyw
Carrier (n ~ 26)
149 • 14 2.1 --- 0.8 9.0 • 3.7 1.33 • 0.47 0 0 0
Mild spherocytosis (n = 30)
128 4.7 27.2 2.44
• 15 +_ 1.9 _+ 14.1 • 0.63 2
911 3
Moderate spherocytosis (n = 49)
108 14.2 41.1 6.28
_+ 17 • 4.7 • 16.0 • 2.26 10 20 22
Severe spherocytosis ~ (n = 2)
60-80 l 1.4 63.0 ND 1 RT 1
ND, Not determinedbecause of regular transfusions;RT, regular transfusions. *Data before transfusion. tValues are expressed as mean -+ SD. :~Normalmean +- SD: 1 • 0.25, w of patients. IIOnlyone crisis per patient.
Table II. Inheritance pattern in four different classes of spherocytosis Dominant
Trait* (n = 26) Mild spherocytosis (n = 30) Moderate spherocytosis (n = 49) Severe spherocytosis (n = 2)
Recessive
Spontaneous
Not k n o w n
0
24
0
2
22
5
1
2
36
7
4
2
0
2
0
0
*Note that carriers with the trait are always related to patients with recessivelyinheritedspherocytosis gility, autohemolysis, and spectrin deficiency in relation to the severity of spherocytosis has been done in a few studies, none of them comprehensive. 3, 8-10 In addition, recommendations about which groups of patients should have splenectomy differ widely. Some authors argue for splenectomy in nearly all cases, 11-t3 but because of the prevalence of postsplenectomy infections, 14, t5 such a broad indication cannot be accepted. Quantitative laboratory criteria indicating the need for splenectomy have not been laid down. We have attempted to develop such criteria on the basis of clinical history, hematologic data, autohemolysis, and osmotic fragility for a sample of patients with spherocytosis, some of whom had spectrin concentration determinations.
METHODS Eighty patients (45 female patients) with the clinical history of spherocytosis were evaluated; 63 patients were children ranging from 4 weeks of age to 15 years, and 17 were adults. All patients were referred to our hospital from 1974 to 1988 because of hemolytic anemia. The diagnosis of spherocytosis was based on the increased osmotic fragility and increased autohemolysis of the erythrocytes, as well
as on the appearance of spherocytes in the peripheral blood smear. The spleen was enlarged to a varying degree. The result of a direct Coombs test was negative in all cases. In addition to the 80 patients, we investigated 27 healthy relatives who had minor signs of spherocytosis such as slightly increased reticulocyte count, osmotic fragility, and autohemolysis. The control sample consisted of healthy children and adults with no signs of increased hemolysis; their ages ranged from 5 to 30 years. Hematology. Hematologic data were determined with a Microcell Counter Sysmex CC-180 (Toa Medical Electronics, Kobe, Japan). The reticulocyte production index was calculated according to Koepke and Koepke.16 The osmotic fragility of the erythrocytes was measured in fresh blood and after 24 hours of incubation, as described by Parpart et al. 17 For a semiquantitative evaluation, the sodium chloride concentrations (gm/dl) at the onset of osmotic hemolysis (1% hemolysis), at the median (50% hemolysis), and at completion (100% hemolysis) were determined. Normal ranges ( + 2 SD) were as previously published. 18 The osmotic fragility in fresh or incubated blood 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
Volume 117 Number 3
Severity of hereditary spherocytosis
411
Fig. t. E•icacy•fsplenect•myinpatientswithmi•d•m•derate•andseverespher•cyt•sis•Hem•gl•binc•ncentrati•n (A) and reticulocyte counts (B) are given before (hatched bars) and after (stippled bars) splenectomy. Mean value and 1 SD range are indicated. HS, Hereditary spherocytosis.
Table III. Spectrin and lipid phosphorus in spherocytosis Spectrin molecules (• 103/cell) Normal (n = 13) Trait (n = 4) Mild spherocytosis (n = 7) Moderate spherocytosis* (n = 6) Severe spherocytosis* (n = 2)
226 _+ 54 244 _+ 77 229 _+ 65"-[ 162 _+ 46 (72%) 142 (62%)
A
Lipid phosphorus (10 -16 mol/cell)
Spectrin per lipid phosphorus (mmol/mol)
3.3 _+ 0.4 3.0 _+ 0.3 3.1 _+ 0.6
1.14 _+ 0.29 1.28 _+ 0.26 1.26 _+ 0.37
3.0 +_ 0.4
0.89 _+ 0.20 (78%) 0.8 (70%)
3.0
Values (except for severe spherocytosis) are expressed as mean +_ SD. Bracket indicates signficance level of c~ = 0.1. *Percentage of the normal value is given in parentheses.
osmotic fragility was described as distinctly increased. Autohemolysis was measured without and with the addition of glucose according to Young et al. 19 The reduction of autohemolysis (percentage of correctability) by the addition of glucose was calculated according to the following formula: t Correctability (%) = 100 X
1-
Autohemolysis with glucose (%) ] Autohemolysis ] without glucose (%),/
Statistical analysis. Statistical evaluations were performed with the Wilcoxon-Mann-Whitney test. 2~ Concentration of speetrin and lipid phosphorus in erythrocytes. Erythrocytes were washed in isotonic sodium chloride solution and pressed through cotton wool to remove leukocytes. For the preparation of lysed erythrocytes, up to 0.5 ml
packed cells (5 X 109 cells) were lysed in the same volume of 0.05 m o l / L tris-(hydroxymethyl)-aminomethane-HC1 buffer, pH 8.2, containing 0.38 m o l / L glycerol, 10% wt/vol sodium dodecyl-sulfate, and 10% vol/vol mercaptoethanol. The samples were incubated at 30 ~ C for 15 minutes, diluted with double distilled water (1 "5), and heated at 100 ~ C for 7 minutes. After centrifugation (10 minutes at 48,000 g) the supernatant was diluted 1:2 with sample buffer. Polyacrylamide slab gel electrophoresis in the presence of sodium dodecyl sulfate of lysed erythrocytes, together with spectrin standards of known concentration, was performed according to Fairbanks et al.21; gels were stained with Coomassie brilliant blue G 250. After being destained, the spectrin bands were cut out. The dye was extracted in 25% pyridine and qiaantitated spectrophotometrically. 22 Spectrin concentration was calculated as copies of a~3 dimer per lysed erythrocyte. For measurement of lipid phosPhorus in
4 12
Eber, Armbrust, and Schrg)ter
The Journal of Pediatrics September 1990
Table IV. Osmotic fragility in spherocytosis Fresh blood (gm NaCl/dl) Osmotic hemolysis (%)
1
50
Incubated blood at 24 hr (gm NaCl/dl) 100
I
50
Normal (n = 13) 0.53 + 0.02 0.43 _+ 0.02 0.17 _+ 0.03]]-] 0.75 _+ 0.05 0.46 _+ 0.03 Trait (n = 26) 0.53 _+ 0.04---I 0.45 _+ 0.04 0.33 + 0.08-a[[ 0.73 _+ 0.09--] 0.55 _+ 0.03 Mild spherocytosis 0.57 + 0'06~------a/ 0.46 _+ 0.03 0.36 + 0.06-a[ 0.88 _+ 0.06--1 0.62 _+ 0.06 (n = 20) / 1 Moderate spherocytosis 0.69 + 0.09---1 0.49 _+ 0.04 0.38 _+ 0.07 ~ - 0.98 _+ 0.08 0.67 _+ 0.08 (n = 30) Severe spherocytosis 0.67 0.49 0.40 0.92 0.59 (n = 2) Values (except for severe spherocytosis)are expressed as mean _+standard deviation.Brackets indicatesignificantdifferences (a = 0.05).
erythrocytes, cholesterol and phospholipids from up to 1 ml packed cells (10 l~ cells) were extracted according to the method of Rose and Oklander. 23 Lipid extracts were ashed at 300 ~ C, and the lipid phosphorus concentration per erythrocyte was determined by the ammonium molybdate color reaction. 24 Spectrin density was defined as the concentration of moles of spectrin per mole of lipid phosphorus. RESULTS
Classification of the patients according to clinical severity. The patients could be separated into three classes of differing clinical severity (mild, moderate, and severe); a fourth class consisted of symptom-free carriers. The criteria for the classification were (1) the concentration of hemoglobin; (2) signs of increased hemolysis (reticulocyte counts, reticulocyte production index, serum concentration of bilirubin); (3) the frequency of severe neonatal hyperbilirubinemia requiring exchange transfusion; and (4) the number of transfusion-dependent aplastic or hemolytic crises (Table I). Heterozygous carriers. Heterozygous carriers were detected exclusively by family screening of patients with symptomatic sphcrocytosis. None of the carriers had any clinical symptoms. Only by considering all available data (rcticulocyte count, concentration of haptoglobin, osmotic fragility, and autohemolysis) could the presence of a genetic abnormality be demonstrated; sphcrocytes were not found. Mild spherocytosis. Patients with mild spherocytosis had a compensated hemolytic anemia. Rcticulocyte counts wcrc 6%, and erythropoiesis was stimulated more than threefold. The serum bilirubin concentration usually exceeded 34 m m o l / L (2 mg/dl), and 93% of the patients had spherocytes. A postnatal exchange transfusion was necessary in 20% of the cases. About half of the patients had had several hemolytic crises and one aplastic crisis, and had required several transfusions. Four patients had gallstones, which were removed by choleeystectomy at the time of the splenectomy. Splenectomy was performed on about half of the patients. Severe spherocytosis. Two unrelated patients had a severe form requiring regular transfusions from birth to sustain a hemoglobin level greater than 60 gm/L. The parents of both patients were carriers of spherocytosis (mild reticulocytosis, slightly elevated osmotic fragility, and raised autohemolysis). Thus these two patients had recessively inherited spherocytosis. One of them already had gallstones at the age of 5 years. Inheritance pattern. The inheritance patterns are given in Table II. The patients belonged to 60 unrelated families. Forty families (66%) had a dominant mode and 12 families (20%) a recessive mode of inheritance. In the latter families, both parents were symptom-free carriers. In four families a spontaneous mutation was inferred, and in another four families no genealogic studies could be performed. Response to splenectomy. Splenectomy was effective, leading to a normal hemoglobin concentration in all patients (Fig. 1). Many patients with moderate spheroeytosis had mild reticulocytosis, indicative of a continuing slightly increased hemolysis after splenectomy. With rare exceptions the serum bilirubin concentration became normal after splenectomy in all patients with mild and moderate spherocytosis. The patient with severe spherocytosis had mild reticulocytosis and a slightly increased bilirubin concentration (19 ~ m o l / L = 1.1 mg/dl) after splenectomy. On average, the three patients with mild spherocytosis had
Volume 117 Number 3
Severity o f hereditary spherocytosis
4 13
,= 33
A
B n=22
28-
9024 ~.
n =22
~ 0
20-
E O r o
16-
