International Journal of Laboratory Hematology

The Official journal of the International Society for Laboratory Hematology

ORIGINAL ARTICLE INTERNAT IONAL JOURNAL OF LABORATO RY HEMATO LOGY

ICSH guidelines for the laboratory diagnosis of nonimmune

hereditary red cell membrane disorders
M.-J. KING*, L. GARC ßON † , J. D. HOYER ‡ , A. IOLASCON § , V. PICARD ¶ , G. STEWART**, P. BIANCHI † † ,
‡‡,1
S.-H. LEE , A. ZANELLA † † , FOR THE INTERNATIONAL COUNCIL FOR STANDARDIZATION IN
HAEMATOLOGY

*Membrane Biochemistry, NHS S U M M A RY

Blood and Transplant, Bristol,
UK Introduction: Hereditary spherocytosis (HS), hereditary elliptocytosis
†
Laboratoire d’H
ematologie, (HE), and hereditary stomatocytosis (HSt) are inherited red cell dis-
Centre de Biologie Humaine,
CHU d’Amiens, Amiens, France orders caused by defects in various membrane proteins. The hetero-
‡
Department of Laboratory geneous clinical presentation, biochemical and genetic
Medicine and Pathology, Mayo abnormalities in HS and HE have been well documented. The need
Clinic Rochester, Rochester,
to raise the awareness of HSt, albeit its much lower prevalence
MN, USA
§
Department of Molecular than HS, is due to the undesirable outcome of splenectomy in these
Medicine & Medical patients.
Biotechnologies, University Methods: The scope of this guideline is to identify the characteristic
Federico II of Naples, Naples,
clinical features, the red cell parameters (including red cell mor-
Italy
¶
H
ematologie Biologique, phology) for these red cell disorders associated, respectively, with
Bic^etre et Facult
e de Pharmacie, defective cytoskeleton (HS and HE) and abnormal cation perme-
AP-HP H^ opital, Universit
e Paris- ability in the lipid bilayer (HSt) of the red cell. The current
Sud, Le Kremlin Bic^etre, France
**Division of Medicine,
screening tests for HS are described, and their limitations are
University College London, highlighted.
London, UK Results: An appropriate diagnosis can often be made when the
††
Hematology Unit, screening test result(s) is reviewed together with the patient’s clini-
Physiopathology of Anemias
Unit, Foundation IRCCS C
a cal/family history, blood count results, reticulocyte count, red cell
Granada Ospedale Maggiore morphology, and chemistry results. SDS–polyacrylamide gel elec-
Policlinico, Milan, Italy trophoresis of erythrocyte membrane proteins, monovalent cation
‡‡
Haematology Department, St.
flux measurement, and molecular analysis of membrane protein
George Hospital, SEALS Central,
Sydney, NSW, Australia genes are specialist tests for further investigation.
Conclusion: Specialist tests provide additional evidence in supporting
Correspondence: the diagnosis and that will facilitate the management of the
May-Jean King, Membrane patient. In the case of a patient’s clinical phenotype being more
Biochemistry, NHS Blood and
Transplant, 500 North Bristol severe than the affected members within the immediate family,
Park, Northway, Filton, Bristol molecular testing of all family members is useful for confirming the
BS34 7QH, UK. diagnosis and allows an insight into the molecular basis of the
Tel.: +44 (0) 117 921 7601;
abnormality such as a recessive mode of inheritance or a de novo
Fax: +44 (0) 117 912 5782;
E-mail: [Link]@nhsbt. mutation.
[Link]

304 © 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 305

1
ICSH Representative.
Table of Contents
doi:10.1111/ijlh.12335 1. Introduction
1.1. Red blood cell membrane structure
Received 4 November 2014; 2. Hereditary spherocytosis
accepted for publication 22 3. Hereditary elliptocytosis and hereditary pyropoikilocytosis
January 2015 4. Hereditary stomatocytosis
5. Coexistence of membranopathy and other red cell disorders
6. Diagnostic tests for red cell membrane defects
Keywords
6.1. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis
Hereditary spherocytosis, hered-
of red cell membrane proteins
itary elliptocytosis, hereditary
6.1.1. Membrane protein defects in HS
stomatocytosis, red cell
6.1.2. Membrane protein defects in HE and HPP
membrane defects, inherited
6.2. Screening tests for hereditary spherocytosis
hemolytic anemia
6.2.1. The Osmotic Fragility Test
6.2.2. The Acid Glycerol Lysis Time Test
6.2.3. Osmotic gradient ektacytometry
6.2.4. The Eosin-50 -Maleimide (EMA) Binding test
7. Molecular genetic analysis
8. Recommendations for laboratory diagnosis of red cell membrane
Disorders
9. Acknowledgements
10. Conflict of Interest
11. References
12. Supporting information (Appendices) online version
Appendix A. Blood smears of different red cell disorders
Appendix B. Screening tests for hereditary spherocytosis
Appendix C. Identification of specific mutations in membrane
protein genes

enriched our basic knowledge in the structural orga-

1. INTRODUCTION
nization of human red cell membrane. In the last
The designation for each of the hereditary red cell 20 years, further progress has been made in under-
membrane disorders denotes the characteristic red cell standing the pathophysiology of a group of rare
morphology expected to be found on the peripheral membrane disorders caused by protein defects in the
blood smear of the patient. Abnormal red cell mor- lipid bilayer, the hereditary stomatocytoses (HSt).
phology is often an indicator but not necessarily spe- The volumes of red cells in certain types of HSt can
cific for the named red cell disorder (Figure S1 in be markedly altered due to abnormal monovalent
Appendix A). In light of new blood cell parameters cation fluxes [2, 3]. Recently, a genome-wide associ-
being measured by the blood cell analyzers [1], it is ation study of hemoglobin concentration and related
possible to use some of the red cell parameters as indi- parameters (Hb, MCH, MCHC, MCV, PCV, and
cators for an initial consideration of membranopathy. RBC) has revealed possible effects of 75 independent
The adoption of SDS–polyacrylamide gel electro- genetic loci on the genetic mechanisms and biologi-
phoresis and molecular genetic techniques since the cal pathways that control red blood cell formation
1970s has made significant inroads into understand- and function [4].
ing the pathophysiology of hereditary spherocytosis Two published Guidelines have already dealt with
(HS), hereditary elliptocytosis (HE), and hereditary the diagnosis and management of hereditary sphero-
pyropoikilocytosis (HPP). Knowing the underlying cytosis [5, 6]. Although hereditary stomatocytosis
causes of these red cell membrane disorders has also (HSt) is about 30- to 40-fold less common than HS, its

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
306 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

inclusion in this guideline is to raise the awareness for phospholipid leaflets. Embedded within this lipid
the identification and differentiation between HS, bilayer are numerous transmembrane proteins and
overhydrated and dehydrated HSt. This is due to an discrete lipid raft domains where GPI-anchored pro-
increased risk of thromboembolic events in those teins reside. This lipid bilayer acts as a barrier for the
patients with these two types of HSt postsplenectomy retention of cations and anions within the red cells
[7, 8]. Variable degrees of amelioration of anemia while it allows water molecules to pass through freely.
postsplenectomy have also been reported [9–11]. Human red cell has high intracellular K+ and low
The writing of this guideline has made use of evi- intracellular Na+ contents when compared with the
dence-based materials in peer-reviewed publications corresponding ion concentrations in plasma. The
from online literature search using key words relevant maintenance of this cation gradient between the red
to the subject matter. The unpublished data presented cell and its environment involves a passive outward
herein are merely for illustration purpose. It is to the movement of K+, which is pumped back by the action
best knowledge of the contributors that the informa- of an ATP-dependent Na+/K+ pump in exchange for
tion presented is accurate. Omissions of other citations Na+ ions. An imbalance of intracellular cation content
were unavoidable due to massive amounts of publica- can result in alteration of red cell volume (e.g. dehy-
tions on these topics. Presented in the Appendices of dration or swelling), which is the underlying cause of
this Guideline are supplementary materials, such as hereditary stomatocytosis, in which the leak rate of
blood smears (Appendix A), background information cation exceeds the pump rate [12].
on selected screening tests (Appendix B), and a selec- Covering about 65% of the cytoplasmic cell surface
tion of mutations associated with the protein genes of the lipid bilayer is a network of spectrin and actin,
(Appendix C). which form a stable ternary complex with protein 4.1
[13]. The contact points between the lipid bilayer and
the spectrin-based network are through the band
1.1. Red blood cell membrane structure
3-ankyrin-protein 4.2 complex and the glycophorin
The human erythrocyte membrane is a laminated C-P55-protein 4.1R complex (Figure 1). The entire
structure comprising of two distinct layers with no framework maintains the shape, integrity, and de-
direct contact with each other (Figure 1). Their com- formability of the red cell. When there is either a
positions and physiological functions are very differ- qualitative or a quantitative defect in any one of the
ent. The outer layer is composed of two asymmetric structural proteins [14], a shortened life span of the

Cation content Glycophorin

Cell mmol/L cells Hereditary Elliptocytosis C/D
[K+] 100
[Na+] 8 Hereditary Spherocytosis
Plasma mmol/L
[K+] 3.5 -5.3 RhAG
[Na+] 133-146

Baseline Carbonic
diffusion anhydrase
Na+ CO2
K+ Hb
Band 3 Glut1
K+ (anion transporter)

Cl-
ATP-dependent
PIEZO1 HCO 3-
Na+/K+ pump
Na+
K+ Cl- Hereditary Stomatocytoses
K+ Cl- co-transporter

Figure 1. Structural organization and functions of various red cell membrane proteins in health and disease.

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 307

circulating red cells due to red cell fragility is mani- A preliminary laboratory diagnosis includes exami-
fested as hemolytic anemia. Abnormal red cell mor- nation of red cell morphology and full blood count
phology is often the first clue to this type of red cells results. Freshly collected blood samples should be
disorders, including hereditary spherocytosis (HS), used for preparation of blood smear to minimize arti-
hereditary elliptocytosis (HE), and hereditary pyropoi- facts. Spherocytes and hyperdense cells (% Hyper
kilocytosis (HPP) [15]. with Hb > 410 g/L) are a positive predictor for HS
when the patient has both a family history and red
cells indices consistent with HS [33], especially when
2 . H E R E D I TA RY S P H E R O C Y TO S I S
≥ 4% Hyper are shown in the CBC. In general,
Hereditary spherocytosis has been reported world- increased % Hyper indicates red cell dehydration [31].
wide, but the highest prevalence is found among the Therefore, they are also present in other hemolytic
northern European populations (estimated 1: 2000 conditions, such as infantile pyknocytosis, acquired
births). The uncoupling of an unstable cytoskeleton immune spherocytosis, DHSt, and congenital dysery-
from the lipid bilayer in localized areas on the red cell thropoietic anemia (CDA) type II. Among the red cell
membrane can hasten red cell vesiculation. Thus, the indices, MCV is not discriminative for HS. MCHC has
resulting spherocytes have a decreased surface/volume been found useful as indicator for HS with concurrent
ratio, dehydrated with reduced deformability. The neonatal jaundice [17]. In the case of HS with com-
shortened life span of HS red cells reflects the intrasp- pensated hemolytic anemia, MCHC is often in the
lenic destruction of red cells due to their sequestration upper limit of the normal range or increased [33].
within the splenic cords [16]. Both the MCV and the mean sphered cell volume
Typical HS patients present with anemia, jaundice, (MSCV) values for red cells are displayed by the Beck-
splenomegaly, and reticulocytosis (Table 1). The dis- man Coulter cell analyzer. The MSCV value is
ease severity of HS is classified on the basis of the expected to be greater than that of MCV as a result of
degree of anemia, namely asymptomatic state, mild, cell volume increase after spherization in a hypo-
moderate, and severe [24]. Most of the patients have osmolar solution during sample analysis. Under the
mild HS, and up to 20–30% have a purely compensated same condition, the MSCV value obtained for HS red
hemolysis due to a balance between reticulocyte pro- cells remains lower than that for the MCV value
duction and red cell destruction. In this category, HS because HS red cells are unable to increase their vol-
could be evoked when finding gallstone, persistent ume further in a hypo-osmolar medium. Thus, a delta
jaundice, clinical splenomegaly, or a transient anemia (MCV-MSCV) value > 9.6 fL is a good indicator for
in a hemolytic crisis during infection or pregnancy HS (90.6% specificity and 100% sensitivity) [34]
(Table 1). The clinical picture for a family with HS is when there is a negative DAT result.
often fairly homogeneous. Variable clinical severity in Automated reticulocyte indices could be useful to
isolated HS families (or kindred) has been linked to co- differentiate hereditary from acquired spherocytosis
inheritance of low expression polymorphism in trans to [35, 36]. As membrane re-modeling begins at the
a HS (Sp) allele, namely recessive HS and the a-spectrin reticulocyte stage in HS, a reduced reticulocyte volume
polymorphism SpaLEPRA allele (Low Expression PRA- under 100 fL is classically observed in HS, except in
gue) [25, 26], which has a frequency of 3.6% in the the neonates [37]. A two-step algorithm is established
normal population [25, 27], and a low expression band for identifying spherocytosis using the reticulocyte
3 allele (Allele Genas) in one family [28]. indices and the detection of hypochromic cells by the
Diagnosis of HS can be difficult in neonates as Sysmex analyzer [35]. The high predictive value for
spherocytic cells are often present after birth. Most of HS (100% sensitivity) relies on the use of reticulocyte
the HS newborns have a normal hemoglobin level at indices to identify a population with a low percentage
birth but develop a transient and sometimes severe of immature reticulocyte fraction, i.e., with a high
anemia during the first few weeks of life, due to the RET/IRF ratio (Rule #1). This ratio was particularly
inability of the infants to mount an appropriate reticu- high (> 19) in mild or subclinical HS. Rule #2 for mod-
locyte response to compensate for the filtering func- erate and severe HS is based on a high percentage of
tion of the spleen [31, 32]. microcytic cells (MicroR) and on a low proportion of

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
308 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

Table 1. Criteria for diagnosis of hereditary spherocytosis and general observations

Mode of inheritance: dominant in 75% cases, and 25% recessive or nondominant HS
Presentation and observation
A. Neonatal period/Infancy
• Neonatal jaundice: common in a few days of life. Risk of kernicterus for those with spherocytosis and
requiring transfusion [17, 18]
• Hb level: normal at birth. Rapid fall in Hb levels within the 1st month after birth
• Anemia: mostly improves during the 1st year of life as high HbF level causes neonatal spherocytosis
B. Childhood/Adulthood
• Persistent jaundice, anemia, splenomegaly, gallstones
• Hemolysis: can be compensated in adults
• Sudden onset of acute anemia: aplastic crisis (Parvovirus B19); acute increase of hemolysis due to viral
infections (e.g. CMV, Herpes 6, EBV)
• Pregnancy in mild HS: aggravated anemia and return to baseline after delivery [19]
C. Elderly: diagnosed in the 7–9th decade in life, often postinfection [20]
D. Chromosome 8p11 translocation or deletion (close to ankyrin gene): spherocytosis and dysmorphic or neurologic
features [21–23]
E. Fatal hydrops fetalis: rare, more likely to happen with a family history of HS [29, 30]
Red cell indices, morphology, and other parameters
• Hemolysis indicators: low or no haptoglobin; raised LDH; hyperbilirubinemia (unconjugated)
• Red cell morphology: Predominantly spherocytes. Mild HS shows occasional spherocytes or only during
hemolytic crisis. Mushroom-shaped cells (association with band 3 defects) (Appendix A). Anisocytosis in
patients with severe hemolytic anemia
Red cell indices:
• Hb: variable, often lower than the normal range. May be normal
• Increased % hyperdense cells on Advia analyzer [32]
• MCHC in the high range or increased, reflecting RBC dehydration [17]
• Increased delta (MCV-MSCV) value on Beckman Coulter analyzers [34, 36]
• Reticulocyte count: increased. Low percentage of immature reticulocyte fraction (increased RET/IRF% ratio)
on Sysmex analyzers, especially in mild HS [35]. Reduced reticulocyte volume under 100 fL, except in
neonates [37]
Exclusions:
• Acquired spherocytosis: autoimmune hemolytic anemia or ABO incompatibility (DAT positive)
• Patients with CDAIIoften presenting with a nondominant HS-like phenotype
• Gilbert’s syndrome: elevated hyperbilirubinemia: no hemolytic feature except if when co-inherited with HS
(increased risk of gallstones)
• Infantile pyknocytosis: may require top-up transfusion in 6–9 months after birth. Resolved with no further
intervention [43]
• Occasional spherocytes in conditions unrelated to membranopathy: Pyrimidine-50 -nucleotidase deficiency
(also shows basophilic stippling) [44]; G6PD deficiency in crisis; PK deficiency or unstable hemoglobin

hypochromic erythrocytes (Hypo-He) among the mi- Although their etiology is different, in several cases,
crocytes (i.e. an increased MicroR/Hypo-He ratio) [35]. such patients were splenectomized based on this HS
Some patients with CDAII are known to present diagnosis without complete clinical remission post-
with mild hemolytic anemia, splenomegaly, and mi- splenectomy. The differentiation between HS and CDA
crospherocytes that give increased osmotic fragility II depends on a correct assessment of red cell morphol-
[38]. These cases could have been initially diagnosed ogy on peripheral blood smear (poikilocytosis) and
to have HS and re-assigned as CDAII [39, 40]. reticulocyte count which can also be compared against

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 309

the Hb level [41]. A reticulocyte count of < 150 9 109/ because most patients either have a compensated
L is within the 90 percentile of the reticulocyte distri- anemia or are asymptomatic. Red cell morphology
bution in a majority of patients with CDAII [41, 42]. (Figure S2 in Appendix A), Hb, and MCV are key
This value falls within the range of 120 9 109– indicators when assessing a patient’s clinical severity.
180 9 109/L for suspected HS cases that would require The variable clinical severity observed in patients
consideration for differentiation from CDAII. with HE is the consequence of either a membrane
protein deficiency or a qualitative defect, respectively,
in protein 4.1R and the spectrin variants. Both can
3 . H E R E D I TA RY E L L I P TO C Y TO S I S A N D
effect an impairment of protein–protein interaction
H E R E D I TA RY P Y R O P O I K I L O C Y TO S I S
within the cytoskeleton. The mild HE clinical picture
Hereditary elliptocytosis (HE) is heterogeneous in presented by a partial 4.1 deficiency can arise from a
respect to the mode of inheritance (dominant to variety of mutational events [52, 53], which are dif-
recessive), clinical severity (compensated anemia to ferent to those mechanisms producing the 4.1 null
transfusion dependent) (Table 2), biochemical (Mem- phenotype [54, 55].
brane protein defects in HE and HPP), and molecular The black populations have a higher incidence of
defects (Table 5 and Table S1 in Appendix C). Finding spectrin defects, whereas protein 4.1 abnormalities are
a small percentage of elliptocytes does not necessarily prevalent among Caucasians. Patients with HE with
rule out HE [51] as this varies between 10% and spectrin defects tend to present a more severe clinical
100%. The diagnosis of HE is incidental (e.g. ellipto- phenotype [56] than those having a partial 4.1 defi-
cytes flagged up by blood counter or postinfection) ciency. This can be related to the severity of the spec-

Table 2. Characteristic features of hereditary elliptocytosis

1. Ethnicity: higher prevalence in countries in the malaria-endemic regions: probable protection against malarial
infection [45]. 1 in 100 births in some African countries [46] and 1: 5000 births among Caucasians. Also found in the
populations in the Mediterranean countries, the Middle East, Japan [47], the Indian subcontinent, and the Far East.
2. Classification based on clinical severity
• Asymptomatic carrier state, nonhemolytic HE: normal Hb level and normal reticulocyte count
• HE with varying degrees of hemolysis: normal Hb but raised reticulocyte count in mild hemolysis; reduced
Hb and raised reticulocyte counts in severe hemolysis
• HE with transient infantile poikilocytosis: transfusion dependent up to the first year, and evolving quickly
into the milder phenotype of the HE parent
• Hereditary pyropoikilocytosis (HPP): severe transfusion-dependent hemolytic anemia, starting in the
neonatal period with Hb of 60–100 g/L and poikilocytosis
• Fatal hydrops fetalis: only one case due to homozygosity in b-spectrin mutation [48] note: Asymptomatic
or mild elliptocytosis can be exacerbated by an infection (e.g. CMV)
3. Red cell morphology
• Elliptocytes present in asymptomatic and nonhemolytic HE
• Hemolytic HE: increasing severity in hemolysis results in red cell fragmentation, showing poikilocytosis due
to a decreased membrane mechanical stability
• HPP: severe microspherocytosis, micropoikilocytosis. MCV between 50 and
4. Exclusions:
• Iron deficiency anemia, thalassemia, and other causes of acquired elliptocytosis, such as burns patients,
myelodysplasia [49], and blood specimens inadvertently kept at high temperature
• Blood group Leach phenotype (Ge-2,-3,-4) in the blood group Gerbich (Ge) system: Individuals with
Gerbich null phenotype are hematologically well and healthy. Their elliptocytic RBCs lack both GPC/D and
20% reduction of protein 4.1 [50]

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
310 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

trin self-association impairment (see Membrane pro- potassium (K+) [8, 60], resulting in an altered hydra-
tein defects in HE and HPP). Hereditary pyropoikilocy- tion status shown by a significant change in mean cell
tosis (HPP) is one form of severe HE which exhibits volume (MCV) [3]. Four subtypes have been identi-
in vivo red cell fragmentation, especially prominent fied: overhydrated HSt, dehydrated HSt, cryohydrocy-
during the first year of life. The percentage of micro- tosis (CHC), and familial pseudohyperkalemia (FP)
cytes (mainly dehydrated mature red cells) is the best (Table 3). Only FP is an asymptomatic trait, whereas
indicator of the severity of the hemolytic anemia [57]. the other three conditions vary in both clinical sever-
The marked microspherocytosis in HPP is caused by a ity and in qualitative features. All subtypes are
decrease of a-spectrin content in these red cells but inherited as dominant, but new mutations are not
normal spectrin in HE red cells. Furthermore, HPP is a uncommon.
biallelic disorder. One of the modifying factor is a low Overhydrated HSt (OMIM 185000, also known as
expression allele SpaLELY (Low Expression LYon) [58, hydrocytosis) is the prototypical example, character-
59], which has a frequency of 20% in all ethnic ized by a moderate-to-severe hemolytic anemia with
groups studied. Unlike HS that asymptomatic parents a mild macrocytosis and often numerous typical sto-
can produce offspring with typical HS, in a majority of matocytes (Figure S3 in Appendix A). The intracellu-
HE/HPP, at least one parent is expected to have a HE lar Na+ and K+ levels are very abnormal, probably
phenotype. due to an increased pore size of the mutated Rh-
associated glycoprotein (RhAG) that facilitates a more
rapid rate of cation leak through the red cell mem-
4 . H E R E D I TA RY S TO M ATO C Y TO S I S
brane [65].
The hereditary stomatocytoses (HSt) are a group of The incidence of dehydrated HSt (OMIM 194380,
hemolytic conditions in which the ‘primary’ lesion is also known as xerocytosis) is the highest in the
a ‘leak’ to the monovalent cations sodium (Na+) and stomatocytosis group (Table 3). There are variable

Table 3. Summary features for subtypes of hereditary stomatocytosis

Overhydrated Dehydrated Cryohydrocytosis Familial Pseudo-

HSt (OHSt) HSt (DHSt) (CHC) hyperkalemia (FP)

Prevalence 1 in million 1 in 10 000 Rare Rare

births births
Morphology Macrocytosis, Stomatocytes Stomatocytes Normal
stomatocytosis Target cells
Hb 80–100 g/L 120–150 g/L 100–120 g/L Normal range
MCV 120–140 fL 98–120 fL Normal Normal or high
MCHC 240–280 g/L 350–370 g/L Normal Normal
Reticulocytosis 10–15% About 10% About 8% Normal
Haptoglobin Undetectable Undetectable Undetectable Present
Intracellular About 409 normal Abnormal, more [K+] leak at low High plasma [K+]
cations Na+/K+ transport rate subtle than temperature and when blood specimen
n OHSt at 4 °C left at RT for several
hours
Glycolytic Reduced 2,3 DPG oxidized Reduced 2,3 Not known Normal
metabolites glutathione, creatine and DPG
ergothioneine [61]

Exclusions:
• Altered RBC cation transport: (a) subtype of HS with a partial band 3 deficiency [62], (b) Band 3 Campinas
[63]: a splicing mutation.
• SAO and CHC have indistinguishable cation permeability defects [64].

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 311

presentations of DHSt: with and without pseudohy- als, these sterols are rejected at the gut. The presence
perkalemia, and perinatal edema [66]. Recently, DHSt of these sterols in the plasma causes a mild hemolytic
is linked to several mutations of PIEZO1 protein gene state with very marked stomatocytosis with no detect-
(FAM38A) [67, 68]. This mechanosensitive channel able cation leak. The key factor which differentiates
protein probably regulates the response of ion fluxes these cases from the cation-leaky stomatocytoses is the
accompanying mechanical stress when circulating macrothrombocytopenia (MPV about 10–15 fL), which
through narrow passages in capillaries [69]. The is presumably caused by the presence in the plasma of
mutated PIEZO1 proteins can cause a gain-of-function these abnormal sterols. The sterols can be measured by
(or increased permeability) [70], and that seems to mass spectrometry of the patient’s plasma.
explain why a healthy athlete with a firm diagnosis of
DHSt presented with hemolysis after every training
5 . C O E X I S T E N C E O F M E M B R A N O PAT H Y A N D
session [71].
OT H E R R E D C E L L D I S O R D E R S
‘Cryohydrocytosis,’ a name coined to describe red
cells which become wet at refrigerator temperatures, A persistently unexplained hemolysis with or without
comes in three forms. The key abnormality is that bizarre red cell morphology (anisocytosis and poikilo-
the cells leak Na+ and K+ profusely at refrigerator cytosis together with a MCV below 70 fL) is often the
temperatures (4–8 °C) [72]. Hemolysis is variable. The cause of attention for further investigation. Taking
low-temperature leaks cause ‘pseudohyperkalemia,’ detailed family history reveals heterogeneous clinical
artifactually high plasma K+ readings due to loss of presentations among the affected parent(s) and sib-
this cation from the abnormal red cells when they are lings. When faced with the proband’s confounding
cooled down from body to room temperatures after clinical presentation and blood smears (Figure S2 in
venesection. The mildest form of cryohydrocytosis Appendix A), the best approach is to undertake a fam-
presents with pseudohyperkalemia only; the next ily study to determine the segregation pattern of two
most severe presents with a mild hemolytic state, or more concurrent red cell disorders.
normocytic red cells, some stomatocytes, and normal The incidence of HE is about 2% in different
MCHC. The most severe form caused by mutations in regions of Africa [46] while thalassemia [77] and
GLUT1 [73] presents with marked hemolysis, marked South East Asian ovalocytosis (SAO) (incidence of 1–
pseudohyperkalemia, and a pediatric neurological syn- 1.5%) are prevalent in South East Asia and the Mela-
drome which is due to deficiency in glucose transport nesia. SAO (caused by a deletion of nine amino acids
across the blood–brain barrier (‘GLUT1 deficiency’). in Band 3) can be a hemolytic condition during early
‘Familial pseudohyperkalemia’ denotes mild hemo- childhood [78] but hematologically normal when
lytic conditions which present with artifactually high reaching adulthood. The co-presence of heterozygous
plasma K+ due to a red cell leak of potassium. Mild a-thalassemia and SAO does not seem to produce
cryohydrocytosis is an example but there are others. more severe hemolysis and clinical symptoms. Co-
The hematology shows minimal abnormalities. The inheritance of SAO with distal renal tubular acidosis
pseudohyperkalemia can be quite misleading. Fludro- (dRTA) is well documented for different ethnic groups
cortisone can be given for presumed adrenal insuffi- in South East Asia. Other reported compound hetero-
ciency. In a French family in which hypertension was zygotes of dRTA (with 25–60% ovalocytes on blood
present [74], a diagnosis of familial hyperkalemic smears) are associated with HbE, a+ thalassemia, and
hypertension (Gordon’s syndrome) [75] was made Hb H disease, respectively [79]. Autosomal dominant
before the artifactual nature of the hyperkalemia was and autosomal recessive dRTA due to band 3 gene
recognized. (SLC4A1) mutations are found in distinct geographic
Stomatocytosis and hemolysis are also encountered locations [80]. The former is localized mainly in the
in phytosterolemia (also known as sitosterolemia) temperate zone while the latter is exclusively in the
[76], a rare recessively inherited metabolic condition tropical regions encompassing Oman in the west to
in which those steroid alcohols made by oily plants the Papua New Guinea in the east [80]. Three cases of
(sitosterol, iso-fucasterol, others) are absorbed in severe HS with dRTA (without SAO) have been
unlimited quantities from the gut. In normal individu- reported [30, 81, 82].

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
312 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

6 . D I AG N O S T I C T E S T S F O R R E D C E L L 6.1.1. Membrane protein defects in HS

MEMBRANE DEFECTS
Confirmation of HS is often based on detecting a
reduction of spectrin (a and b spectrin), ankyrin, band
6.1 Sodium dodecyl sulfate-polyacrylamide gel
3, or protein 4.2 (Figure 2). A specific membrane pro-
electrophoresis of red cell membrane proteins
tein deficiency can represent either a direct quantita-
SDS–polyacrylamide gel electrophoresis (PAGE) is the tive effect due to a mutation in that protein gene or an
standard technique used not only for the quantitation outcome secondary to the primary mutational event in
of relative membrane protein contents [e.g. decreased a different membrane protein gene. For instance, a
proteins in both HS (Figure 2a) and HE, and increased primary ankyrin gene (ANK1) mutation can cause a
spectrin dimer in HE/HPP (Figure 3a)], but it can also secondary partial a-spectrin reduction or a combined
reveal the abnormal proteins present in the red cells, deficiency of ankyrin/spectrin or spectrin/protein 4.2.
such as CDAII band 3 (Figure 2a) and spectrin vari- An isolated partial protein 4.2 deficiency can result
ants in HE/HPP (Figure 3b). Except for the visualiza- from either an EPB42 (protein 4.2 gene) mutation or a
tion of band 3 (25% of total red cell protein) and SLC4A1 (band 3 gene) mutation by which the protein
stomatin (band 7.2b in Figure 2a), this technique can 4.2 binding site on the mutated Band 3 is abrogated.
not detect GLUT1, RhAG, and Piezo1 for hereditary The highest number of mutations linking to HS resides
stomatocytosis. in the primary genes SPTB, ANK1, and SLC4A1 [15],

(a)

P4.1
P4.2

HS P4.2 (-) HE(P4.1 )

(b) Milan Naples

Bristol + spleen Splenectomy Families Patients

n 333 259 41 220 580
Spectrin a39% 31% 41% 15% 16%
Ankyrin b6% 3% 10% 60% 56%
Band 3c 22% 54% 46% 17% 19%
P4.2 11% 1% 0% 0.4% 0.2%
Normal protein 8% 11% 3% 8% 9%
Sp/P4.2 11% n/a n/a n/a n/a

Figure 2. Membrane protein deficiencies in HS (a) Laemmli and Fairbanks buffer systems are used for gradient gels.
Band 2 in Laemmli gel (4–16% linear gradient) is separated out into b spectrin and ankyrin in the Fairbanks gel
(3.5–17% exponential gradient). (b) Membrane protein deficiencies identified in three laboratories: Bristol
(unpublished data), Milan [83], and Naples [84]. a: the spectrin deficient group comprised of 31% a-spectrin
deficiency only and 8% combined ab-spectrin deficiency. b: Ankyrin deficiency included Ankyrin only (n = 7),
combined Ank/P4.2 (n = 10), Ank/Sp (n = 7), and Ank/Sp/P4.2 (n = 7). c: Band 3 deficiency included both isolated
band 3 and a combined B3/P4.2 reduction (for the three laboratories).

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 313

(a) (b)

(c)

Figure 3. Spectrin analysis for confirmation of HE/HPP and the locations of spectrin variants (a) Dimer content for
HE/HPP; (b) Spectrin variants for HE/HPP from the domain I of a-spectrin (Mr 80 kDa) after trypsin digestion on
ice for 20 h [86], a-spectrin is cleaved into five domains (aI to aV) and b spectrin into four domains (I to IV). The
numbers for the spectrin variants indicate the Mr of each tryptic peptide; aI/36, aI/46–50 (lane 3, including a 25-kDa
peptide), aI/65 (lane 1: homozygote; lane 2: heterozygote), aI/74 (lane 4), and aI/78; N: normal control. (c) Locations
of spectrin variants on the a-spectrin repeats (a1 to a22 repeats) and the location of SpaLELY allele. Note: Most of a
and b chains are composed of linear repeats (i.e. homologous triple helical segments). Each repeat is about 106
amino acids in length and is organized into a triple helix. The arrangement is that helix 2 and 3 in one repeat lock
into helix 1 of the next repeat. The self-association site (circle with dotted line) comprises the N-terminus of a
chain (mainly the aI domain, Mr 80 kDa) and the C-terminus of the b chain. Mutations in SpaLELY allele affect
repeats a18 and a21. Mutations in the a-spectrin and b-spectrin (SPTB) genes can result in producing the same aI/
74
variant. The SpaI/74 variant from mutations in exon 30 of SPTB has been found mainly among Caucasian [90]
and reported in a Filipino family [91]. (cf. Table S1 in Appendix C)

and their corresponding membrane protein deficien- deficiency (the Milan data) [83]. This finding is con-
cies are shown in Figure 2b. sistent with the masking of ankyrin deficiency in HS
A higher incidence of partial spectrin deficiency is by reticulocytosis [85] due to higher ankyrin content
associated with moderate and severe HS. Since only a in young red cells than in aged red cells.
handful of pathogenic SPTA mutations have been
reported, this observation can be attributed to the
6.1.2. Membrane protein defects in HE and HPP
effect of the cis inheritance of the SpaLEPRA allele
either homozygous or heterozygous for this low One-dimensional SDS-PAGE can detect both protein
expression polymorphism [25–27]. There is no defini- 4.1R deficiency and spectrin variants. The latter are
tive correlation between clinical phenotype and the same for both patients with HE and patients with
specific membrane proteins although a spectrin defi- HPP (Figure 3). The severe hemolysis in HPP is
ciency was found more frequently in children than in attributed to higher spectrin dimer content [56] with
adults at presentation [83]. a concomitant reduction in a-spectrin (Figure 3a). By
Between 8 and 11% of patients with HS gave with contrast, it is normal spectrin content in asymptom-
no detectable reductions in membrane proteins (Fig- atic HE. Limited trypsin digestion of spectrin reveals
ure 2b). Splenectomy seemed to have improved the the defective region in the spectrin protein as a tryp-
sensitivity of SDS-PAGE in detecting a partial ankyrin sin-resistant peptide, which is often localized to the

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
314 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

domain I of a spectrin (designated as aI/80: Mr also give a positive result (i.e. increased red cell lysis).
80 000) [86] (Figure 3b). The defects identified These include immune hemolytic anemia, recent
within this domain correlate with the extent of blood transfusion (i.e. lysis of recently transfused
impairment in the spectrin self-association site, and RBCs ex vivo due to depletion of ATP in these cells),
the severity of hemolytic anemia. This conclusion is RBC enzyme deficiencies (e.g. G6PD and pyruvate
based on two findings: (i) The proximity of the a- kinase deficiencies), and unstable hemoglobin vari-
spectrin defect (or variant) to the self-association site ants. The OF test result has to be interpreted together
of a and b chains (Figure 3c) and (ii) The proportion with family history and examination of the peripheral
of the mutant HE spectrin a chain recruited into the blood smear.
cytoskeleton, which, in turn, depends on the pres- A flow cytometric osmotic fragility test is available,
ence in trans (or in cis) of the low expression SpaLELY which is based on a greater susceptibility of HS red
allele [58, 59]. cells to lyse in a medium spiked with deionized water.
Identification of spectrin variants from other Although it shows high specificity and sensitivity for
domains requires two-dimensional gel electrophoresis HS [94], it is not known whether this method can
of tryptic spectrin digest. The detection of SpaII/21 also detect red cells with abnormal membrane perme-
[87], SpaII/31[88], or Spectrin St. Claude [89] in the ability to cation.
homozygous probands is easier than in the heterozy-
gous family members due to the elevated spectrin
6.2.2. The acid glycerol lysis time test
dimer content found only in the homozygotes. Since
the locations of these variants are further away from The original glycerol lysis time (GLT) test [95] mea-
the spectrin self-association site (Figure 3c), the sures the time taken for 50% hemolysis of a blood
increased dimer content solely in homozygotes indi- sample in a buffered solution of hypotonic saline/glyc-
cates a long-range effect of a spectrin mutation on erol. Glycerol retards the entry of water into the red
self-association site. It could be a conformational cells, thus prolonging the lysis time. Addition of
change in the helix harboring the mutation within 0.0053M sodium phosphate lowered the pH of the
the triple helical repeat unit [56]. buffered solutions to 6.85 with improvement of the
sensitivity and specificity of this test for HS. This is
the basis of the acidified glycerol lysis time (AGLT)
6.2. Screening tests for hereditary spherocytosis
test [96]. As a result, the lysis of normal red cells can
The screening tests presented in Table 4, except ekta- be measured at a more manageable time of > 900s,
cytometry, are in regular use in general hematology instead of lysis at 23s–45s. The AGLT could also give
laboratories. positive results in acquired spherocytosis, such as
AIHA, in about one-third of the pregnant women,
and in some patients with chronic renal failure and
6.2.1. The osmotic fragility test
with myelodysplastic syndrome [96, 97].
The traditional OF test uses 14 concentrations of NaCl,
ranging from 0.1 to 0.8 g/dL NaCl [92]. However,
6.2.3. Osmotic gradient ektacytometry
modifications are available: a 4-tube method using
four concentrations of NaCl solution (0.5 g/dL for un- Osmotic gradient ektacytometry monitors the extent
incubated cells, 0.60, 0.65, and 0.75 g/dL NaCl for of red cell shape change within a range of osmotic
incubated RBCs) [93], and a 17-tube method with the gradient (i.e. modulation of cellular water content)
last 3 tubes at 0.85, 0.9, and 1.0 g/dL of NaCl (Milan). under a known shear stress (Appendix B.1). An osmo-
Spherocytes have less resistance to lysis at each NaCl tic gradient deformability index (DI) profile provides
concentration when compared to normal RBCs. Pre- the information about cell water content and hetero-
incubation of the whole blood sample for 24 h at geneity of deformability. The advantage of this
37 °C will enhance the degree of cell lysis. technique is the production of distinct deformabili-
The drawback of the OF test is a lack in specificity, ty profiles for several types of red cell disorders
as other congenital red cell defects or conditions can (Figure 4).

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 315

Table 4. Laboratory tests for screening red cell disorders

OF test AGLT test Ektacytometry EMA Binding test
(Mayo and Milan data) (Milan data) (Bicêtre data in Fig 4) (Data from Bicêtre, Bristol, Mayo, and Milan laboratories)
1 2 3
Result for normal Traditional OF: (14 tubes) > 900 sec normal values are defined in Lab Instrument n MCF(SD) Ratio (SD) %Reduction
controls 17 tube assay (Milan) plus 0.85, 0.9 each assay. (SD)
4
and 1.0 g/dL NaCl Ranges: Bristol FC500 120 53.9 (1.9) 1.00 (0.04) *0 %
4-tube OF: % Cell lysis Omin Ohyper DImax Canto II 120 12412 (638) 1.01 (0.05) *0 %
NaCl (g/dL) Male Female 135-155 335-360 > 0.4 Bicêtre5 Navios 391 70.4 (5.2) 1.00 (0.04) 0 % (4%)
0.50 (fresh) 0.0 – 47.8 0.0-31.1 (mosmol/kg) Milan5 Canto II 1445 10830 (2290) 1.00 (0.05) 0% (5%)
Incubated: Mayo FACSCaliburs 213 629.2 (34.4) 1.00 (0.05) * 0%
0.60 18.7 – 67.4 10.9 – 65.5
0.65 4.4 – 36.6 0.2 – 39.3
0.75 0.8 – 9.1 0.0 - 10.9
HS Increased fragility < 900 sec Reduced fluorescence
4-tube assay: positive when lysis ↓ DImax (↓ deformability) Lab Instrument n MCF(SD) Ratio (SD) %Reduction (SD)
greater than the normal ranges for ↑Omin (i.e., ↑ osmotic fragility), Bristol4 FC500 43 42.9 (4.7) 0.78 (0.09 ) * -22%
all 4 NaCl concentrations and/or ↓ Ohyper 5
Canto II 43 9660 (1098) 0.77 (0.08) * -23%
17-tube assay: positive when lysis Positive result when at least two Bicêtre Navios 228 53.5 (8.4) 0.75 (0.11) -25% (-11%)
greater than the normal range or abnormal parameters including Milan5 Canto II 381 7800 (1900) 0.75 (0.1) -27% (-10%)
abnormal results for at least 2 decreased DImax Mayo FACSCaliburs 213 444.3 (34.4) 0.71 (0.05) * -29%
consecutive NaCl concentrations
AIHA Normal to increased fragility Normal to Normal to HS-like DI Normal fluorescence
HS-like
HE Normal Normal ↓ DImax (Trapezoidal profile) Normal or borderline MCF
(non-hemolytic)
HPP Increased fragility ↓ lysis time ↓↓ DImax (↓↓deformability) ↓↓ fluorescence (lower than the HS range)
distorted trapezoidal profile
SAO Reduced fragility No data No deformability ↓ fluorescence (within the range for HS) #
CDA type II Increased fragility for some cases Normal to Isolated ↓ deformability to HS- ↓ fluorescence within the HS range for some, otherwise normal result
HS-like like DI profile (↓DImax, Omin
and O hyper are within the normal
range )
Cryohydrocytosis No data No data No data ↓ fluorescence (within range for HS) ¶
Dehydrated HSt Reduced fragility > 900 sec Normal DImax (deformability), ↓ Normal or ↑ fluorescence
(exclusion of thalassemia) Ohypo and ↓Ohyper; DI profile left
of normal (dehydration)
Overhydrated Increased fragility No data Normal deformability Normal or ↑ fluorescence
HSt (HS-like) DI profile right of normal
deformability (over-hydration)

*MCF: mean channel fluorescence readings are shown.

†
The ratio was obtained by (MCF of test sample/mean MCF of normal controls used).
‡
The % reduction was obtained by (mean MCF of Normal controls used – MCF of Test)/mean MCF of normal controls
used.
§
These data were taken from a prospective study in which the same samples were analyzed contemporaneously using
FC500 and Canto II [102]. The normal controls shown represent a separate group from the normal controls used for
calculating ratios. Six normal controls were used per assay.
¶
Both the Milan and Bic^etre laboratories use % reductions in reporting patient results [42, 108]. The MCF readings are
not used at all.
**The % Reduction shown for Bristol and Mayo was calculated using the following equation: (1- ratio) 9100%. Both
laboratories use MCF in result reporting.
††
Distinct red cell morphology for SAO (Figure S3-L in Appendix A).
‡‡
The amount of Band 3 reduction detected in SDS-PAGE cannot explain the severity of the patient’s hemolytic ane-
mia. An increase of plasma [K+] on storage of patient’s blood sample in the refrigerator or a delay of 5–6 h in sample
arrival at the testing laboratory [72].

For producing consistent DI profiles, same day dration (Figure 4). All blood samples are required to
blood collection and testing are preferred. If necessary, give a negative result in the direct antiglobulin test
overnight storage of blood specimens at 4 °C is still prior to performing ektacytometry. This is due to
acceptable. Red cells stored over 24 h will give a pro- overlapping results obtained for HS and the sphero-
file of overhydration. Antibodies bound to red cells cytes produced in AIHA [37], especially among jaun-
can affect each parameter (DImax, Omin, Ohyper) diced neonates [31].
(Appendix B.1). HE is characterized by a trapezoidal curve with
The DI profiles for HS show a decreased DImax moderately decreased IDmax but normal Omin and
together with an increased Omin (reflecting osmotic Ohyper results. By contrast, the DI profile of HPP red
fragility) and/or decreased Ohyper reflecting cell dehy- cells presents a markedly reduced IDmax (Figure 4).

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
316 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

DI AIHA DI Typical HE DI DI
Normal control Normal control

Osmolality
Osmolality
DHSt Hemoglobin H
Typical HS HPP

Mild HS SE Asian Sickle cell disease βThalassemia

DImax ovalocytosis (SS) trait

Ohyper

Omin Data provided by the bicêtre laboratory

Figure 4. Osmotic deformability profiles of red cells from normal control, AIHA, HS, HE/HPP, DHSt, and
hemoglobinopathies.

SAO red cells are rigid, and they do not deform in can be interpreted as atypical HS. Therefore, SDS-
this assay. The DHSt curve has a normal IDmax, and PAGE is often performed in the absence of family his-
its left-shifted DI profile indicates red cell dehydra- tory of HS for differential diagnosis. If necessary, there
tion (Figure 4). Although the DI profiles for DHSt will be a follow-up appointment for family members
and sickle cell disease look similar (Figure 4), ektacy- and/or repeat testing of the proband after several
tometry is currently the only simple and reliable (> 3) months.
screening test for the diagnosis of DHSt. Laser-assisted optical rotational red cell analyzer
Exclusion of hemoglobinopathy and iron deficiency (LORCA) presumably works in a principle similar to
should be considered when an abnormal DI profile is ektacytometry although the way of generating the de-
obtained. A ‘flat-top’ DI profile can be produced by formability index is different. An evaluation of both
either thalassemia (Figure 4) or iron deficiency. This ektacytometry and LORCA by using red cells from a
kind of profile for thalassemia red cells is consistent wider range of disorders (not just for HS, HE, overhy-
with resistance to osmotic lysis. drated and dehydrated HSt) and clinical condition(s)
The differentiation between HS and DHSt is (e.g. hyposplenism [98]) will establish whether abnor-
straightforward due to the left shift of the DHSt curve. mal deformability profiles are exclusive to those red
It can be challenging to distinguish between HS and cell disorders associated with defective intrinsic mem-
CDAII using ektacytometry as some profiles for CDAII brane proteins.

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 317

features namely, chronic mild hemolytic anemia (Hb

6.2.4. The Eosin-50 -Maleimide (EMA) binding test
80–100 g/L), raised bilirubin, gallstones, reticulocyto-
The EMA binding test [99] can detect HS red cells as sis, and splenomegaly. By inference, using the EMA
well as SAO, cryohydrocytosis, and some cases of binding test together with a second screening test can
CDA type II [42, 99]. Their differentiation from HS probably detect all patients with HS [42, 104].
[39, 40, 100] can be made using red cell morphology Table 4 shows the reference ranges for normal con-
(Appendix A) and SDS-PAGE (Figure 2a). Further- trols and patients with HS, which give the different
more, this technique can also detect HPP red cells formats of result presentation. Furthermore, these val-
(Appendix B.2) that give mean channel fluorescence ues can serve as a guide for other users when planning
(MCF) readings about 25% lower than that for HS to establish this technique in a laboratory. However,
[101] and high MCF results for the red cells with high these ranges from three laboratories (Bic^ etre, Milan,
MCV [102]. The EMA binding test can be used alone and Mayo) have not taken into consideration of a gray
when adequate clinical and laboratory information on area for HS (Figure 5). Interpretation of ‘gray area’
the patient are available [6]. The alternative approach results requires the consideration of family history,
is to use different combinations of screening tests, clinical presentation with evidence of a hemolytic
namely the OF test/the EMA binding test [6], AGLT/ anemia as well as carrying out a family study to
the EMA binding test [42], and ektacytometry/the include proband, both parents and siblings [102]. A
EMA binding test [103]. trend for HS can be identified based on results from
As a general rule, whenever the MCF reading is the mildly affected family members [105].
within the range for typical HS, a membrane protein The drawback of the EMA binding test is a lack of
deficiency is usually detected in the patient’s red cells universal reference ranges for normal controls and HS
(Figure 5). Exceptions to this were found in about as they are currently established by individual labora-
0.9% of patients (four in 403 cases) suspected with HS tories due to the different scales for mean fluorescence
and unexplained hemolytic anemia (an audit in intensity (MFI) shown on flow cytometers from differ-
Bristol) to have given normal EMA results. Yet, their ent manufacturers (Table 4). Harmonization of result
subsequent SDS-PAGE results showed a membrane presentation can facilitate comparison of patient
protein deficiency: reduced a-spectrin in three patients results between laboratories [106]. This issue together
and one was band 3 deficient. All three spectrin defi- with appropriate cutoff value for distinguishing
cient patients had presented with typical clinical between HS and normal controls was already raised
by several testing laboratories [102, 107–109].

70

7 . M O L E C U L A R G E N E T I C A N A LYS I S
Mean channel fluorescence (MCF)

60
Table 5 shows the wide genetic heterogeneity of
50
hereditary spherocytosis, HE/HPP, and hereditary
Gray
45.5 area stomatocytosis as defects in some of the membrane
40
proteins can cause more than one type of red cell dis-
orders. Consideration for molecular analysis is a case-
30 by-case decision, which is often the last option in a
diagnostic plan for the patient. A general approach is
20 that molecular analysis of membrane protein genes
Normal Sp/4.2 B3/4.2 Ank/4.2 Sp/Ank/4.2 Normal
Adults Sp P4.2 Ank Sp/Ank B3 protein for HS and HE does not add extra information for the
n = 168 132 35 36 13 7 10 7 7 59 27
patient whose family is already known to have the
red cell disorder. A recessive mode of inheritance, and
Figure 5. Box plot showing the MCF readings from a case of suspected de novo mutation or compound
the EMA binding test for patients with HS classified heterozygosity do warrant further investigation using
based on membrane protein deficiencies
DNA sequencing.
(unpublished data from Bristol).

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
318 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

Table 5. Genes involved in red cell membrane defects

Protein (Gene) Chromosomal location Exons (number) mRNA length (bp) Disease

Alpha spectrin (SPTA1) 1q22-23 52 7999 HS, HE

Beta-spectrin (SPTB) 14q23-24.1 35 10 063 HS, HE
Ankyrin (ANK1) 8p11.2 43 8297 HS
Band 3 (SLC4A1) 17q21-22 20 4965 HS, SAO, CHC dRTA
Protein 4.1R (EPB41) 1p33-34.2 21 5930 HE
Protein 4.2 (EPB42) 15q15-21 13 4553 HS
RhAG (RHAG) 6p12-21 10 1912 OHSt
PIEZO1 (FAM38A) 16q24.2-16qter 51 1912 DHSt
Glut1 (SLC2A1) 1p31.3-35 10 3318 CHC
ABCB6 (ABCB6) 2q35-36 19 3016 FP

ABCB6- ATP binding cassette (ABC) transporter in the B subfamily, also known as mitochondrial porphyrin trans-
porter, which imports porphyrin into mitochondria during heme biosynthesis [120]. dRTA, distal renal tubular acidosis;
HS, hereditary spherocytosis; HE, hereditary elliptocytosis; SAO, South East Asian ovalocytosis, DHSt, dehydrated
hereditary stomatocytosis; OHSt, overhydrated hereditary stomatocytosis, CHC, cryohydrocytosis; FP, familial pseudo-
hyperkalemia.
Websites for protein and gene sequences: DNA Data Bank of Japan [Link]
EMBL European Bioinformatics Institute [Link]
National Center for Biotechnology Information [Link]

Molecular assessment of the presence of several SDS-PAGE can detect the typical hypoglycosylated
modifiers genes, such as UGT1A1 promoter polymor- CDAII band 3 (Figure 2a) if there is uncertainty
phisms and mutations in the HFE gene, could assist in between HS and CDAII for a patient. The alternative
the prediction of the course of disease development option is to perform molecular analysis of the SEC23B
[18]. If available, such analysis can confirm the diag- gene for confirmation of CDA II [112, 113].
nosis of HS for patients with apparently normal par- In some cases of HE that have a proband with a
ents. These patients could have either a genuine markedly severe clinical presentation and a milder HE
recessive pattern of inheritance (i.e. homozygosity or phenotype in both parents, demonstration of a hypo-
double heterozygosity for an a-spectrin mutation) or morphic allele (such as spectrin aLELY allele) co-inher-
an apparently recessive one caused by de novo muta- ited in-trans to a SpHE allele (i.e. a structural spectrin
tion of ankyrin or b-spectin genes (Table S1 in Appen- defect) in the proband will explain the clinical sever-
dix C). To discriminate between various possibilities, ity [58, 59]. The molecular basis for a protein 4.1R
SDS-PAGE can be performed prior to quantitation of abnormality can be caused by one of the mutational
ankyrin or b spectrin gene expression (e.g. the events: gene deletions, decreased 4.1mRNA levels
reverse-transcribed amplified cDNA from the region of caused by promoter mutations or unstable mRNA [53,
several polymorphisms). A de novo monoallelic expres- 54], or a decreased protein synthesis caused by trans-
sion of these genes is often the underlying cause of lational defects [55]. SDS-PAGE can detect the protein
recessive inheritance pattern in patients with HS 4.1 deficiency.
[110]. However, a reduction of ankyrin could be the Reaching a preliminary diagnosis of hereditary
result of a mutation in the SLC4A1 or ANK1. stomatocytosis takes time due to a lack of specific dis-
Finally, a spectrin alleles in patients with isolated ease indicators for testing in general laboratory. The
spectrin deficiency and a co-inheritance of defective b basic information is an increase of MCV and MCHC,
spectrin gene have to be investigated at the molecular time- and temperature-dependent K+ leak from red
level to distinguish between a mutation or a polymor- cells, stomatocytes, or target cells (Table 3). Using ekt-
phic allele (e.g. spectrin aLEPRA) responsible for the acytometry is the best option for OHSt and DHSt,
recessive pattern of inheritance [25, 26, 111]. albeit only available in a very few testing centers in

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 319

Europe. If hereditary stomatocytosis is suspected at mutations in nonmembrane protein genes that can
presentation of a patient, confirmation of the diagno- have a modifying effect on the clinical phenotype.
sis can be carried out by means of sequencing the
appropriate candidate gene (Table 5). For instance,
sequencing of the coding region for the C-terminal 8 . R E C O M M E N DAT I O N S F O R L A B O R ATO RY
domain of SLC4A1 will be able to confirm cyrohydro- D I AG N O S I S O F R E D C E L L M E M B R A N E
cytosis as nine amino acid substitutions in the region DISORDERS
between aa-687 and aa-796 of band 3 protein have
Aim to reduce the turnaround time due to request for
been reported for cryohydrocytosis [114].
unnecessary additional tests; e.g. SDS-PAGE of eryth-
Recent advances in the next generation sequencing
rocyte membrane proteins and molecular testing:
(NGS) technologies have transformed the genetics
study of human diseases [115]. This is an era of • A testing laboratory can make an informed decision
unprecedented productivity. Exome sequencing, the when a test requisition form gives adequate patient
targeted sequencing of the protein-coding portion of information and the reason for the test(s) required.
the human genome, has been shown to be a powerful Useful information includes a tentative diagnosis, the
and cost-effective method for detection of disease patient’s clinical presentation, a DAT-negative result,
variants underlying Mendelian disorders. There are reticulocyte count, and markers for hemolysis (Fig-
several new attempts to use this technology to diag- ure 6), and family history.
nosing in one reaction all the possible causes of red • State clearly if the reason for test request is for
cell cytoskeleton defects. This method could be cer- exclusion of membrane defect.
tainly useful mainly for autosomal dominant cases. • Send a pretransfusion blood specimen or wait for
However for autosomal recessive inheritance, the lack complete clearance of residual transfused red cells
of capability of NGS to diagnose mutations in intronic before sending a blood specimen to testing laboratory.
sequence as well in regulatory sequence and deletion- It takes 8 weeks for complete clearance of transfused
al forms could impair this methodology [116, 117]. cells from circulation.
Application of NGS in the diagnosis of red cell
For hereditary spherocytosis:
membrane disorders has recently been reported [118,
119]. Presented herein is a brief summary on the • Spherocytes on blood smear has a good positive pre-
experience of the Naples laboratory in their validation dictive value for HS in a patient with a family history
of various molecular techniques using cases of known and compatible red cells indices. In the absence of a
specific membrane protein defects. After identifying family history, DAT and a screening test for HS should
the underlying protein gene mutation in a proband both be performed.
using the NGS, other family members (e.g. parents • Signs for mild HS: occasional spherocytes in the
and affected relatives) will always be tested by other peripheral blood smear, together with borderline re-
less expensive molecular techniques (namely PCR and ticulocytosis, slightly abnormal biochemical markers
sequencing) to confirm the inherited mutation(s). In a of hemolysis with or without splenomegaly. These
patient, more than one membrane protein gene can observations may be diagnostic of mild HS after
be identified to have a mutation. Verification of the exclusion of other causes of spherocytosis namely,
gene mutation directly responsible for the clinical pregnancy, autoimmune hemolysis, CDAII, systemic
phenotype is based on functional assay using the hematological diseases, and some enzyme deficien-
Xenopus oocyte expression system. As a general esti- cies.
mate, a wide genome search can identify causative • All the screening tests presented can detect typical HS.
mutation(s) in about 60% of cases. The correct assign- If in doubt, use two screening tests (Table 4). Equally
ment rate (i.e. matching the diagnosis with the mem- important is an awareness of when a screening test can
brane protein gene) increases to 80% with whole give false-positive results with other red cell disorders
exome sequencing. A 90% assignment rate is feasible and false-negative results with isolated HS cases.
when using targeted sequencing. The added advantage • If available, molecular testing is considered when all
of targeted sequencing is the concomitant detection of screening tests give negative or equivocal results for a

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
320 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

Figure 6. Diagnostic flow chart for HS, HE, and HSt. 1Exclusion of CDAII when test results are within in HS but the
patient’s clinical presentation is not typical of HS. Exclusion of hemoglobinopathy (e.g. unstable Hb) when the test
result is mildly HS-like for a patient whose MCV is < 65 fL and the blood smear shows anisocytosis and
poikilocytic red cells. 2: exclusion of HS.

proband presenting with characteristic features of mem- For hereditary stomatocytosis:

brane disorder, whereas the family members are hema-
tologically well and healthy. Identifying the molecular
• Important to make a firm diagnosis as splenectomy
is not beneficial to a patient with dehydrated HSt, and
basis of the hemolytic anemia facilitates genetic counsel-
overhydrated HSt [7, 121]. The confirmatory tests
ing or when considering for splenectomy.
include ion flux measurement, ektacytometry, or
• Detection of SpaLEPRA allele by PCR technique (cf. LORCA for overhydrated HSt, dehydrated HSt, and
Table S1 in Appendix C) can explain the more severe
DNA sequencing for specific genes (Table 5).
clinical phenotype in a proband than in other HS fam-
ily members.
• If none of the above is available, use screening tests
for differential diagnosis of HS and other rarer red cell
For hereditary elliptocytosis: disorders (Table 4). If available, use flame photometry
to quantitate plasma K+, intracellular [K+] and [Na+]
• In the absence of a family history of HE and only a
in fresh and stored specimens from patient and a
few elliptocytes on the blood smear, it is advised to
paired normal control being kept at room temperature
confirm the diagnosis of HE for the proband [51].
and in the refrigerator [72].
• If available, carry out SDS-PAGE for quantitation of
protein 4.1 and spectrin analysis (spectrin dimer con-
• If a patient does not respond to splenectomy (which
was performed based on a diagnosis of HS) [9], re-
tent and spectrin variant) (Figure 3). Or, use ektacy-
examine the patient for the likelihood of HSt.
tometry/LORCA to obtain characteristic DI profile
(Figure 4).
• In HPP, detection of SpaLELY allele in trans to a SpHE 9 . AC K N OW L E D G E M E N T S
allele by PCR technique can explain the severe clinical
The Hematology Departments in the 1Foundation
phenotype in a majority of patients. New mutational
IRCCS Ca’ Granda Ospedale Maggiore of Milan
events can also cause HPP-like phenotype (cf. Table
(Italy), 2CHU Bic^
etre (France), and the 3Mayo Clinic
S1 in Appendix C).

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 321

Rochester (USA) formed a Harmonization Group. The er, 3Adam J. Wood, MD, 3Jennifer L. Oliveira, MD,
purpose was to retrieve their respective archival data and 3Michelle E. Savedra for their assistance in this
for the EMA binding test carried out using normal work.
controls and patients with HS in the last 3–6 years.
Statistical analysis was carried out individually to
10. CONFLICT OF INTEREST
establish the reference ranges for normal controls
and patients with HS. We thank 1Elisa Fermo PhD, All the contributors of this Guideline have declared
2
Caroline Mayeur-Rousse MD, 2Aurore Gerstenmey- no conflict of interest.

11. REFERENCES potassium red cells. New Eng J Med spherocytosis gene on chromosome 8.
1968;278:573–81. Ann Int Med 1983;99:192–3.
1. Brugnara C, Mohandas N. Red cell indi- 11. Glader BE, Fortier N, Albala MM, Nathan 23. Chilcote RR, Le Beau MM, Dampier C,
ces in classification and treatment of ane- DG. Congenital hemolytic anemia associ- Pergament E, Verlinsky Y, Mohandas N,
mias: from M.M. Wintrobes’s original ated with dehydrated erythrocytes and Frischer H, Rowley JD. Association of red
1934 classification to the third milleni- increased potassium loss. New Eng J Med cell spherocytosis with deletion of short
um. Curr Opin Hematol 2013;20:222–30. 1974;291:491–6. arm of chromosome 8. Blood 1987;69:
2. Stewart GW, Turner EJ. The hereditary 12. Flatt J, Bruce LJ. The hereditary stom- 156–9.
stomatocytoses and allied disorders: con- atocytosis. Haematologica 2009;94:1039– 24. Eber SW, Armbrust R, Schroter W. Vari-
genital disorders of erythrocyte mem- 41. able clinical severity of hereditary sphe-
brane permeability to Na and K. 13. Salomao M, Zhang X, Yang Y, Lee S, rocytosis: relation to erythrocytic spectrin
Baillieres Clin Haematol 1999;12:707–27. Hartwig JH, Chasis JA, Mohandas N, An concentration, osmotic fragility, and
3. Gallagher PG. Disorders of red cell vol- X. Protein 4.1R-dependent multiprotein autohemolysis. J Pediatrics 1990;117:
ume regulation. Curr Opin Hematol complex: new insights into the structural 409–16.
2013;20:201–7. organization of the red blood cell mem- 25. Wichterle H, Hanspal M, Palek J, Jarolim
4. Van der Harst P, Zhang W, Leach IM, brane. Proc Nat Acad Sci USA 2008;105: P. Combination of two mutant alpha
Rendon A, et al. Seventy-five genetic loci 8026–31. spectrin alleles underlies a severe sphero-
influencing the human red blood cell. 14. Giorgi M, Cianci CD, Gallagher G, Mor- cytic hemolytic anemia. J Clin Invest
Nature 2012;492:369–75. row JS. Spectrin oligomerization is coop- 1996;98:2300–7.
5. Guitton C, Garcßon L, Cynober T, Gauthier eratively coupled to membrane assembly: 26. Dhermy D, Steen-Johnsen J, Bournier O,
F, Tchernia G, Delaunay J, Leblanc T, a linkage targeted by many hemolytic Hietet G, Cynober T, Tchernia G, Grand-
Thuret I, Bader-Meunier B. Sph
erocytose anemias? Exp Mol Pathol 2001;70:215– champ B. Coinheritance of two a-spec-
h
er
editaire: recommendations pour le 30. trin gene defects in a recessive
diagnostic et la prise en charge chez 15. Delaunay J. The molecular basis of spherocytosis family. Clin Lab Haem
l’enfant (Hereditary spherocytosis: hereditary red cell membrane disorders. 2000;22:329–36.
guidelines for the diagnosis and manage- Blood Rev 2007;21:1–20. 27. Boivin P, Galand C, Devaux L, Lecomte
ment in children). Arch P
ediatr 16. Molnar Z, Rappaport H. Fine structure of MC, Garbarz M, Dhermy D. Spectrin
2008;15:1461–73. the red pulp of the spleen in hereditary alpha IIa variant in dominant and non-
6. Bolton-Maggs P, Langer JC, Iolascon A, spherocytosis. Blood 1972;39:81–98. dominant spherocytosis[published erra-
Tittensor P, King M-J. Guidelines for the 17. Christensen RD, Henry E. Hereditary tum appears in Human Genetics 1994
diagnosis and management of hereditary spherocytosis in neonates with hyperbi- May; 93, 614.]. Hum Genet
spherocytosis- 2011 update. Br J Haema- lirubinemia. Pediatrics 2012;125: 1993;92:153–6.
tol 2012;156:37–49. 120–5. 28. Alloisio N, Maillet P, Carr
e G, Texier P,
7. Stewart GW, Amess JAL, Eber SW, 18. Gallagher PG. Abnormalities of the eryth- Vallier A, Baklouti F, Philippe N, Dela-
Kingswood C, Lane PA, Smith BD, Ment- rocyte membrane. Pediatr Clin North Am unay J. Hereditary spherocytosis with
zer WC. Thrombo-embolic disease after 2013;60:1349–62. band 3 deficiency. Association with a
splenectomy for hereditary stomatocyto- 19. Pajor A, Lehoczky D, Szak
acs Z. Preg- non-sense mutation of the band 3 gene
sis. Br J Haematol 1996;93:303–10. nancy and hereditary spherocytosis. (Allel Lyon), and aggravation by a low-
8. Delaunay J. The hereditary stomatocytos- Report of 8 patients and a review. Arch expression allele occurring in trans
es: genetic disorders of the red cell Gynecol Obstet 1993;253:37–42. (Allele Genas). Blood 1996;88:1062–9.
membrane permeability to monovalent 20. Friedman EW, Williams JC, van Hook L. 29. Whitfield CF, Follweiler JB, Lopresti-
cations. Sem Hematol 2004;41:165–72. Hereditary spherocytosis in the elderly. Morrow L, Miller BA. Deficiency of a-
9. Lock SP, Smith RS, Hardisty RM. Stomat- Am J Med 1988;84:513–6. spectrin synthesis in burst-forming units-
ocytosis: a hereditary red cell anomaly 21. Kimberling WJ, Taylor RA, Chapman erythroid in lethal hereditary spherocyto-
associated with haemolytic anaemia. Br J RG, Lubs HA. Linkage and gene localiza- sis. Blood 1991;78:3043–51.
Haematol 1961;7:303–14. tion of hereditary spherocytosis (HS). 30. Ribeiro ML, Alloisio N, Almeida H, Go-
10. Zarkowsky HS, Oski FA, Sha’afi R, Sho- Blood 1978;52:859–67. mes C, Texier P, Lemos C, Mimoso G,
het SB, Nathan DG. Congenital hemo- 22. Bass EB, Smith SW, Stevenson RE, Rosse Morl
e L, Bey-Cabet F, Rudigoz RC, Dela-
lytic anemia with high sodium, low WF. Further evidence for location of the unay J, Tamagini G. Severe hereditary

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
322 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

spherocytosis and distal renal tubular aci- cohort study of 205 cases with congenital brane skeletal protein 4.1 in normal and
dosis associated with the total absence of dyserythropoietic anemia type II: defini- abnormal red blood cells. Sem Hematol
band 3. Blood 2000;96:1602–4. tion of clinical and molecular spectrum 1993;30:58–73.
31. Saada V, Cynober T, Brossard Y, Schis- and identification of new diagnostic 54. Baklouti F, Morinere M, Haj-Kh
elil A,
chmanoff PO, Sender A, Cohen H, Dela- scores. Am J Hematol 2014;89:E169–75. F
en
eant-Thibault M, Gruffat H, Cout
e Y,
unay J, Tchernia G. Incidence of 42. Bianchi P, Fermo E, Vercellati C, Marcel- Ninot A, Guitton C, Delaunay J. Homo-
hereditary spherocytosis in a population lo AP, Porretti L, Cortelezzi A, Barcellini zygous deletion of EP41 genuine AUG-
of jaundiced neonates. Pediatr Hematol W, Zanella A. Diagnostic power of labo- containing exons results in mRNA splic-
Oncol 2006;23:387–97. ratory tests for hereditary spherocytosis: ing defects, NMD inactivation and pro-
32. Delhommeau F, Cynober T, Schischma- a comparison study on 150 patients tein 4.1R complete deficiency in
noff PO, Rohrlich P, Delaunay J, Mohan- grouped according to the molecular and hereditary elliptocytosis. Blood Cells Mol
das N, Tchernia G. Natural history of clinical characteristics. Haematologica Dis 2011;47:158–65.
hereditary spherocytosis during the first 2012;97:516–23. 55. Morinere M, Ribeiro L, Venezia ND, De-
year of life. Blood 2000;95:393–7. 43. Eyssette-Guerreau S, Bader-Meunier B, guillien M, Maillet P, Cynober T, Del-
33. Cynober T, Mohandas N, Tchernia G. Garcßon L, Guitton C, Cynober T. Infantile hommeau F, Almeida H, Tamagnini G,
Red cell abnormalities in hereditary pyknocytosis: a cause of haemolytic Delaunay J, Baklouti F. Elliptocytosis in
spherocytosis: relevance to diagnosis and anaemia of the newborn. Br J Haematol patients with C-terminal domain muta-
understanding of the variable expression 2006;133:439–42. tions of protein 4.1 correlates with
of clinical severity. J Lab Clin Med 44. Bianchi P, Fermo E, Alfinito F, Vercellati encoded messenger RNA levels rather
1996;128:259–69. C, Baserga M, Ferraro F, Guzzo I, Rotoli than with alterations in primary protein
34. Bros
eus J, Visomblain B, Guy J, Mayna- B, Anella A. Molecular characterization structure. Blood 2000;95:1834–41.
di

e M, Girodon F. Evaluation of mean of six unrelated Italian patients affected 56. Delaunay J, Dhermy D. Mutations
sphered corpuscular volume for predict- by pyrimidine-50 -nucleotidase deficiency. involving the spectrin heterodimer con-
ing hereditary spherocytosis. Int J Lab Br J Haematol 2003;122:847–51. tact site: clinical expression and altera-
Hematol 2010;32:519–23. 45. Facer CCA. Erythrocytes carrying muta- tions in specific function. Sem Hematol
35. Mullier F, Lainey E, Fenneteau O, Da tions in spectrin and protein 4.1 show 1993;30:21–33.
Costa L, Schillinger F, Bailly N, Cornet Y, differing sensitivities to invasion by Plas- 57. Silveira P, Cynober T, Dhermy D, Mo-
Chatelain C, Dogne JM, Chatelain B. modium falciparum. Parasitol Res handas D, Tchernia G. Red blood cell
Additional erythrocytic and reticulocytic 1995;81:52–7. abnormalities in hereditary elliptocytosis
parameters helpful for diagnosis of 46. Dhermy D, Carnevale O, Blot I, Zohoun and their relevance to variable clinical
hereditary spherocytosis: results of a I. Hereditary elliptocytosis in Africa. Lan- expression. Am J Clin Pathol
multicentre study. Ann Hematol cet 1989;1:225. 1997;108:391–9.
2011;90:759–68. 47. Yawata Y. Characteristics of red cell 58. Alloisio N, Morl
e L, Mar
echal J, Roux
36. Lazarova E, Pradier O, Cotton F, Gulbis membrane disorders in the Japanese pop- AF, Ducluzeau MT, Guetarni D, Pothier
B. Automated reticulocyte parameters for ulation. Rinsho Byori 1997;45:367–76. B, Baklouti F, Ghanem A, Kastally R,
hereditary spherocytosis screening. Ann 48. Gallagher PG, Weed SA, Tse WT, Benoit Delaunay J. SpaV/41: a common spectrin
Hematol 2014;93:1809–18. L, Morrow JS, Marchesi SL, Mohandas polymorphism at the aIV-aV domain
37. Da Costa L, Mohandas N, Sorette M, N, Forget BG. Recurrent fatal hydrops junction. Relevance to the expression
Grange MJ, Tchernia G, Cynober T. Tem- fetalis associated with a nucleotide substi- level of hereditary elliptocytosis due to a-
poral differences in membrane loss lead tution in the erythrocyte b-spectrin gene. spectrin variants located in trans. J Clin
to distinct reticulocyte features in heredi- J Clin Invest 1995;95:1174–82. Invest 1991;87:2169–2177.
tary spherocytosis and in immune hemo- 49. Alanio-Br
echot C, Schischmanoff PO, 59. Wilmotte R, Mar
echal J, Morl
e L, Baklo-
lytic anemia. Blood 2001;98:2894–9. F
en
eant-Thibaut M, Cynober T, Tchernia uti F, Philippe N, Kastally R, Kotula J,
38. Danise P, Amendola G, Nobili B, Perrotta G, Delaunay J, Garcßon L. Association Delaunay J, Alloisio N. Low expression
S, Miraglia del Giudice E, Matarese SMR, between deficit in myeloid malignancies allele aLELY of red cell spectrin is associ-
Iolascon A, Brugnara C. Flow-cytometric and acquired protein 4.1R deficiency: a ated with mutations in exon 40 (aV/41
analysis of erythrocytes and reticulocytes retrospective analysis of 6 patients. Am J polymorphism) and intron 45 and with
in congenital dyserythropoietic anaemia Hematol 2008;83:275–8. partial skipping of exon 46. J Clin Invest
type II (CDA II): value in differential 50. Telen MJ, Le Van Kim C, Chung A, Car- 1993;91:2091–6.
diagnosis with hereditary spherocytosis. tron JP, Colin Y. Molecular basis for el- 60. Stewart GW. Hemolytic disease due to
Clin Lab Haem 2001;23:7–13. liptocytosis associated with glycophorin C membrane ion channel disorders. Curr
39. Heimpel H, Anselstetter V, Chrobak L, and D deficiency in the Leach pheno- Opin Hematol 2004;11:244–50.
Denecke J, Einsiedler B, Gallmeier K, type. Blood 1991;78:1603–6. 61. Darghouth D, Koehl B, Hellier JF, Mada-
Griesshammer A, Marquardt T, Janka- 51. F
eo C, Dhermy D. Confirming or ruling linski G, Bovee P, Bosman G, Delaunay
Schaub G, Kron M, Kohne E. Congenital out hereditary elliptocytosis. Acta Hae- J, Junot C, Romeo P-H. Alterations of
dyserythropoietic anemia type II: epide- matol 1988;79:116. red blood cell metabolome in overhydrat-
miology, clinical appearance, and prog- 52. Conboy JG, Chasis JA, Winardi R, Tcher- ed hereditary stomatocytosis. Haemato-
nosis based on long-term observation. nia G, Kan YW, Mohandas N. An iso- logica 2011;96:1861–5.
Blood 2003;102:4576–81. form-specific mutation in the protein 4.1 62. Bruce LJ, Robinson H, Guizouarn H,
40. Dukka S, King M-J, Hill Q. The diagnosis gene results in hereditary elliptocytosis Borgese F, Harrison P, King M-J, Goede
challenge of congenital dyserythropoietic and complete deficiency of protein 4.1 in JS, Coles SE, Gore DM, Lutz HU, Ficarel-
anaemia: two cases of ‘CDA type II’. J erythrocyte but not in nonerythroid cells. la R, Layton M, Iolascon A, Ellory JC,
Clin Path 2014;67:367–9. J Clin Invest 1993;91:77–82. Stewart GW. Monovalent cation leaks in
41. Russo R, Gambale A, Langella C, Andolfo 53. Conboy JG. Structure, function, and human red cells are caused by single
I, Unal S, Iolascon A. Retrospective molecular genetics of erythroid mem- amino acid substitutions in the transport

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 323

domain of the band 3 chloride-bicarbon- 72. Haines PG, Jarvis HG, King S, Noo- the anion exchange 1 causing spherocy-
ate exchanger, AE1. Nat Genet rmohamad FH, Chetty M, Fisher J, Hill tosis and distal renal tubular acidosis.
2005;37:1258–63. P, Nicolaou A, Stewart GW. Two further Biochem J 2010;426:379–88.
63. Lima PRM, Gontijo JAR, Lopes de Faria British families with the ‘cryohydrocyto- 82. Sinha R, Agarwal I, Bawazir WM, Bruce
JB, Costa FF, Saad STO. Band 3 campin- sis’ form of hereditary stomatocytosis. Br LJ. Distal renal tubular acidosis with
as: a novel splicing mutation in the band J Haematol 2001;113:932–7. hereditary spherocytosis. Indian Pediatr
3 gene (AE1) associated with hereditary 73. Flatt JF, Guizouarn H, Burton N, Borgese 2013;50:693–5.
spherocytosis, hyperactivity of Na+/Li+ F, Tomlinson RJ, Forsyth RJ, Baldwin 83. Mariani MG, Barcellini W, Vercellati C,
countertransport and an abnormal renal SA, Levinson BE, Quittlt P, Aguilar-Mar- Marcello AP, Fermo E, Pedotti P, Bosch-
bicarbonate handling. Blood 1997;90: tinez P, Delaunay J, Stewart GW, Bruce etti C, Zanella A. Clinical and hematolog-
2810–8. LJ. Stomatin-deficient cryohydrocytosis ic features of 300 patients affected by
64. Guizouarn H, Borgese F, Gabillat N, Har- results from mutations in SLC2A1: a hereditary spherocytosis grouped accord-
rison P, Geode JS, McMahon C, Stewart novel form of GLUT1 deficiency syn- ing to the type of the membrane protein
GW, Bruce L. Southeast Asian ovalocyto- drome. Blood 2011;118:5267–77. defect. Haematologica 2008;93:1310–7.
sis and the cryohydrocytosis form of 74. Beaurain G, Mathieu F, Grootenboer S, 84. Iolascon A, Avvisati RA. Genotype/phe-
hereditary stomatocytosis show virtually Fiquet B, Cynober T, Tchernia G, Dela- notype correlation in hereditary sphero-
indistinguishable cation permeability unay J, Jeunemaitre X. Dehydrated cytosis. Haematologica 2008;93:1283–8.
defects. Br J Haematol 2011;152:655–64. hereditary stomatocytosis mimicking 85. Miraglia del Giudice E, Francese M, Poli-
65. Bruce LJ, Guizouarn H, Burton NM, Gab- familial hyperkalaemic hypertension: to R, Nobili B, Iolascon A, Perrotta S.
illat N, Poole J, Flatt JF, Brady RL, Borgese clinical and genetic investigation. Eur J Apparently normal ankyrin content in
F, Delaunay J, Stewart G. The monovalent Haematol 2007;78:253–9. unsplenectomized hereditary spherocyto-
cation leak in overhydrate stomatocytic 75. Gordon RD. Syndrome of hypertension sis patients with the inactivation of one
red blood cells results from amino acid and hyperkalemia with normal glomeru- ankyrin (ANK1) allele. Haematologica
substitutions in the Rh-associated glyco- lar filtration rate. Hypertension 1997;82:332–3.
protein. Blood 2009;113:1350–7. 1986;8:93–102. 86. Speicher DW, Morrow JS, Knowles WJ,
66. Basu AP, Carey P, Cynober T, Chetty M, 76. Rees D, Iolascon A, Carella M, O’Marca- Marchesi VT. Identification of proteolyti-
Delaunay J, Stewart GW, Richmond S. igh A, Kendra JR, Jowitt SN, Wales JK, cally resistant domains of human eryth-
Dehydrated hereditary stomatocytosis Vora A, Makris M, Manning N, Nicolaou rocyte spectrin. Proc Natl Acad Sci USA
with transient perinatal ascites. Arch Dis A, Fisher J, Mann A, Machin SJ, 1980;77:5673–7.
Child Fetal Neonatal Ed 2003;88:F438–9. Clayton PT, Gasparini P, Stewart GW. 87. Alloisio N, Morl
e L, Pothier B, Roux AF,
67. Zarychanski R, Schulz VP, Houston BL, Stomatocytic haemolysis and macro- Mar
echal J, Ducluzeau MT, Benhadji-
Maksimova Y, Houston DS, Smith B, thrombocytopenia (Mediterranean stom- Zouaoui Z, Delaunay J. Spectrin Oran
Rinehart J, Gallagher PG. Mutations in aotcytosis/macrothrombo-cytopenia) is (aII/21), a new spectrin variant concern-
the mechanotransduction protein PIEZO1 the haematological presentation of phyt- ing the aII domain and causing severe el-
are associated with hereditary xerocyto- osterolaemia. Br J Haematol 2005; liptocytosis in the homozygous state.
sis. Blood 2012;120:1908–15. 130:297–309. Blood 1988;71:1039–47.
68. Andolfo I, Alper SL, De Franceschi L, Au- 77. Fuchareon S, Winichagoon P. Haemoglo- 88. Alloisio N, Wilmotte R, Morl
e L, Baklouti
riemma C, Russo R, De Falco L, Valle- binopathies in Southeast Asia. Indian J F, Mar
echal J, Ducluzeau M-T, Denoroy
fuoco F, Esposito MR, Vandorpe DH, Med Res 2011;134:498–506. L, F
eo C, Forget BG, Kastally R, Dela-
Shmukler BE, Narayan R, Montanaro D, 78. Laosombat V, Viprakasit V, Dissaneevate unay J. Spectrin Jendouba: an aII/31
D’Armiento M, Vetro A, Limongelli I, Zu- S, Leetanaporn R, Chotsampancharoen T, spectrin variant that is associated with el-
ffardi O, Glader BE, Schrier SL, Brugnara Wongchanchailert M, Kodchawan S, liptocytosis and carried a mutation dis-
C, Stewart GW, Delaunay J, Iolascon A. Thongnoppakun W, Duangchu S. Natural tant from the dimer self-association site.
Multiple clinical forms of dehydrated history of Southeast Asian ovalocytosis Blood 1992;80:809–15.
hereditary stomatocytosis arise from during the first 3 years of life. Blood 89. Fournier CM, Nicolas G, Gallagher PG,
mutations in PIEZO1. Blood 2013; Cells Mol Dis 2010;45:29–32. Dhermy D, Grandchamp B, Lecomte MC.
121:3925–35, S1-S12. 79. Yenchitsomanus PT, Kittanakom S, Run- Spectrin St Claude, a splicing mutation of
69. Bae C, Gnanasambandam R, Nicolai C, groj N, Cordat E, Reithmeier RAF. the human a-spectrin gene associated
Schs F, Gottlieb PA. Xerocytosis is caused Molecular mechanisms of autosomal with severe poikilocytic anemia. Blood
by mutations that alter the kinetics of dominant and recessive distal renal tubu- 1997;89:4584–90.
the mechanosensitive channel PIEZO1. lar acidosis caused by SLC4A1 (AE1) 90. Tse WT, Lecomte MC, Costa FF, Garbarz
Proc Natl Acad Sci USA 2013;110:E1162– mutations. J Mol Genet Med 2005;1:49– M, F
eo C, Boivin P, Dhermy D, Forget
8. 62. BG. Point mutation in the beta-spectrin
70. Albuisson J, Murthy SE, Bandell M, Cos- 80. Khositseth S, Bruce LJ, Walsh SB, Bawa- gene associated with alpha I/74 heredi-
te B, Louis-dit-Picard H, Mathur J, zir WM, Ogle GD, Unwin RJ, Thong MK, tary elliptocytosis. Implications for the
F
en
eant-Thibault M, Tertian G, de Jaure- Sinha R, Choo KE, Chartapisak W, King- mechanism of spectrin dimer self-associa-
guiberry JP, Syfuss PY. Dehydrated watanakul P, Sumboonnanonda A, Vas- tion. J Clin Invest 1990;86:909–16.
hereditary stomatocytosis is linked to uvattakul S, Yenchitsomanus P, Wrong 91. Dhermy D, Galand C, Bournier O, King
gain-of-function mutations in mechani- O. Tropical distal renal tubular acidosis: M-J, Cynober T, Roberts I, Kanyike F,
cally activated PIEZO1 ion channels. Nat clinical and epidemiological studies in 78 Adekile A. Co-inheritance of a- and b-
Commun 2013;4:1884. patients. Q J Med 2012;105:861–77. spectrin gene mutations in a case of
71. Platt OS, Lux SE, Nathan DG. Exercise- 81. Chiu C, Woods N, Sawasdee N, Guizo- hereditary elliptocytosis. Blood
induced hemolysis in xerocytosis. Eryth- uarn H, Pellissier B, Borgese F, Yenchit- 1998;92:4481–2.
rocyte dehydration and shear sensitivity. somanus PT, Gowrishankar M, Cordat E. 92. Parpart AK, Lorenz PB, Parpart ER, Gregg
J Clin Invest 1981;68:631–8. Band 3 Edmonton I, a novel mutant of J, Chase AM. The osmotic resistance (fra-

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
324 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS

gility) of human red cells. J Clin Invest 103. Da Costa L, Galimand J, Fenneteau O, 111. Iolascon A, Avvisati RA, Piscopo C.
1947;26:636–40. Mohandas N. Hereditary spherocytosis, Hereditary spherocytosis. Transfus Clin
93. Larson CJ, Scheidt R, Fairbanks VF. The elliptocytosis, and other red cell mem- Biol 2010;17:138–42.
Osmotic Fragility Test for Hereditary brane disorders. Blood Rev 2013;27:167– 112. Schwarz K, Iolascon A, Verissimo F, Tre-
Spherocytosis: use of EDTA-anticoagulat- 78. de N, Horsley W, Chen W, Par BH, Kopf-
ed blood stored at 4° C for up to 104. Crisp RL, Solari L, Vota D, Garcia E, Mi- ner KP, Holzmann K, Russo R, Eposito
96 hours. Am J Clin Pathol 1988;90:507. guez G, Chamorro ME, Schvartzman GA, RM, Spano D, Falco LD, Heinrich K, Jog-
94. Park SH, Park C-J, Lee B-R, Cho Y-U, Alfonso G, Gammella D, Caldarola S, gorst B, Rojewski MT, Perrotta S, De-
Jang SS, Kim NY, Koh KN, Im HJ, Seo Riccheri C, Vittori D, Venegas B, Nesse A, necke J, Pannicke U, Delaunay J,
JJ, Park ES, Lee JW, Yoo KH, Jung HL. Donat H. A prospective study to assess Pepperkok R, Heimpel H. Mutations
Comparison study of the eosin-50 malei- the predictive value for hereditary sphe- affecting the secretory COPII coat protein
mide binding test, flow cytometric osmo- rocytosis using five laboratory tests (cryo- SEC23B cause congenital dyserythropoi-
tic fragility tests, and cryohemolysis test hemolysis test, eosin-50 maleimide flow etic anemia type II. Nat Genet
in the diagnosis of hereditary spherocy- cytometry, osmotic fragility test, auto- 2009;41:936–40.
tosis. Am J Clin Pathol 2014;142: hemolysis test, and SDS-PAGE) on 50 113. Bianchi P, Fermo E, Vercellati C, Bos-
474–84. hereditary spherocytosis families in chetti C, Barcellini W, Lurlo A, Marcel-
95. Gottfried EL, Robertson NA. Glycerol Argentina. Ann Hematol 2011;90:625–34. lo AP, Righetti PG, Zanella A.
lysis time of incubated erythrocytes in 105. King M-J, Bruce LJ, Dhermy D. Hemo- Congenital dyserythropoietic anemia
the diagnosis of hereditary spherocytosis. lytic anemias associated with disorders of type II (CDAII) is caused by mutations
J Lab Clin Med 1974;84:746–51. erythrocyte membrane and cytoskeleton in the SEC23B gene. Hum Mutat
96. Zanella A, Izzo C, Rebulla P, Zanuso F, proteins. In: Laboratory Hematology 2009;30:1292–8.
Perroni L, Sirchia G. Acidified glycerol Practice, Kotte-Marchant K, Davis B. 114. Barneaud-Rocca D, Pellissier B, Borgese
lysis test: a screening test for spherocyto- (eds). Chichester, UK: Wiley Blackwell; F, Guizourn H. Band 3 missense muta-
sis. Br J Haematol 1980;45:481–6. 2010: 602–14. tions and stomatocytosis: insight into the
97. Bucx MJ, Breed WP, Hoffmann JJ. Com- 106. Miller GW, Myers GL, Gantzer ML, Kahn molecular mechanism responsible for
parison of acidified glycerol lysis test, SE, Sch€ onbrunner ER, Thienpont LM, monovalent cation leak. Int J Cell Biol
Pink test and osmotic fragility test in Bunk DM, Christenson RH, Eckfeldt JH, 2011;2011:13682.
hereditary spherocytosis: effect of incuba- Lo SF, N} ubling CM, Sturgeon CM. Road- 115. Solomon BD. Obstacles and opportunities
tion. Eur J Haematol 1988;40:227–31. map for harmonization of clinical labora- for the future of genomic medicine. Mol
98. Prendki V, Ndour PA, Jais X, Ciceron L, tory measurement procedures. Clin Genet Genomic Med 2014;2:205–9.
Settegrna C, Buffect P. Reduced deform- Chem 2011;57:1108–17. 116. Khincha PP, Savage SA. Genomic charac-
ability of circulating erythrocytes: a mar- 107. Girodon F, Garcßon L, Bergoin E, Largier terization of the inherited bone marrow
ker for hyposplenism. Amer J Hematol M, Delaunay J, Feneant-Thibault M, failure syndromes. Semin Hematol
2012;87:E81–2. Maynadi
e M, Couillaud G, Moreira S, Cy- 2013;50:333–47.
99. King MJ, Behrens J, Rogers C, Flynn C, nober T. Usefulness of the eosin-5- malei- 117. Gerrard G, Valga~ no
n M, Foong HE, Kas-
Greenwood D, Chambers K. Rapid flow mide cytometric method as a first-line peraviciute D, Iskander D, Game L,
cytometric test for the diagnosis of mem- screening test for the diagnosis of heredi- M€ uller M, Aitman TJ, Roberts I, de la
brane cytoskeleton-associated haemolytic tary spherocytosis: comparison with ekta- Fuente J, Foroni L, Karadimitris A. Tar-
anaemia. Br J Haematol 2000;111:924– cytometry and protein electrophoresis. Br get enrichment and high-throughput
33. J Haematol 2008;140:468–70. sequencing of 80 ribosomal protein genes
100. CLSI. Red Blood Cell Diagnostic Testing 108. Mayeur-Rousse C, Gentil M, Botton J, to identify mutations associated with Dia-
Using Flow Cytometry, Approved Guide- Thibaut MF, Guitton C, Picard V. Testing mond-Blackfan anaemia. Br J Haematol
line –Second Edition. CLSI document for hereditary spherocytosis; a French 2013;162:530–6.
H52-A2. Wayne, PA: Clinical and Labo- experience. Haematologica 2012;97:e48– 118. Christensen RD, Nussenzveig RH, Yaish
ratory Standards Institute; 2014. Sec- 9. (Bianchi P, Fermo E, Zanella A. Reply HM, Henry E, Eggert LD, Agarwal AM.
tion 8: Red Blood Cell Shape Diseases: to ‘Testing for hereditary spherocytosis: a Causes of hemolysis in neonates with
Cellular Defects and Clinical Implication French experience’ Haematologica 2012; extreme hyperbilirubinemia. J Perinatol
(pp 29-41) and Appendix C: Differential 97(12): e52.) 2014;34:616–9.
Diagnosis using the Eosin-5-Maleimide 109. Mackiewicz G, Bailly F, Favre B, Guy J, 119. Van Wijk R, Vogel M, Elferink M, Nijman
Binding Test (pp 89-93). Maynadi
e M, Girodon F. Flow cytometry I, Cuppen E, van Solinger W, van Amstel
101. King M-J, Telfer P, MacKinnon H, Lan- test for hereditary spherocytosis. Haema- HKP, van der Zwaag B. Next generation
gabeer L, McMahon C, Darbyshire P, tologica 2012;97:e47. (Bianchi P, Fermo sequencing in the diagnosis of red cell
Dhermy D. Using the eosin-5-maleimide P, Zanella A. Reply to ‘Flow cytometry membrane disorders. Int J Lab Hematol
binding test in the differential diagnosis test for hereditary spherocytosis’. Haema- 2014;36(Suppl. 1):18, abstract O45.
of hereditary spherocytosis and heredi- tologica, 2012; 97 (12): e50-51.) 120. Krishnamurthy PC, Du G, Fukuda Y, Sun
tary pyropoikilocytosis. Cytometry B 110. Miraglia del Giudice E, Lombardi C, D, Sampath J, Mercer KE, Wang J, Sosa-
2008;74B:244–50. Francese M, Nobili B, Conte ML, Amen- Pineda B, Murti KG, Schuetz JD. Identifi-
102. Hunt L, Greenwood D, Heimpel H, Noel dola G, Cutillo S, Iolascon A, Perrotta S. cation of a mammalian mitochondrial
N, Whiteway A, King M-J. Toward the Frequently de novo monoallelic expres- porphyrin transporter. Nature 2006;
harmonization of result presentation for sion of b spectrin gene (SPTB) in children 443:586–9.
the eosin-5-maleimide binding test in the with hereditary spherocytosis and iso- 121. Smith BD, Segel GB. Abnormal erythro-
diagnosis of hereditary spherocytosis. lated spectrin deficiency. Br J Haematol cyte endothelial adherence in hereditary
Cytometry B 2015;88B:50–7. 1998;101:251–4. stomatocytosis. Blood 1997;89:3451–6.

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325
 M.-J. KING ET AL. | GUIDELINES FOR RBC MEMBRANE DISORDERS 325

12. Supporting Information Appendix S2 Screening tests for hereditary sphero-

cytosis.
Additional Supporting Information may be found in
Appendix S3 Identification of specific mutations in
the online version of this article:
membrane protein genes.
Appendix S1 Blood smears of different red cell disor-
ders.

© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37, 304–325