ORIGINAL ARTICLE Familial pseudohyperkalemia in blood donors: a novel mutation with implications for transfusion practice Waleed M. Bawazir,1,2 Joanna F. Flatt,1 Jonathan P. Wallis,3 Augusto Rendon,4 Rebecca A. Cardigan,5 Helen V. New,6 Michael Wiltshire,5 Lizanne Page,7 Catherine E. Chapman,8 Gordon W. Stewart,9 and Lesley J. Bruce1
BACKGROUND: Familial pseudohyperkalemia (FP) is a dominantly inherited condition in which red blood cells (RBCs) have an increased cold-induced permeability to monovalent cations. Potassium leaks into the supernatant of all stored blood with time, but FP RBCs leak potassium more rapidly. We investigated two unrelated blood donors whose RBC donations demonstrated unexpectedly high potassium after 5 and 6 days’ storage. We matched the observed pattern of RBC cation leak to a previously recognized family with FP (FP-Cardiff) and investigated the likely cause with targeted DNA analysis. STUDY DESIGN AND METHODS: Cation leakage from the donor RBCs and from standard donor units was measured. DNA analysis of donors and family members with FP-Cardiff was performed. Allele frequencies were obtained from human variation databases. RESULTS: Both implicated donors were found to have increased cold-induced potassium leak identical in pattern to affected members of the family with FP-Cardiff. We found a heterozygous substitution Arg723Gln in the ATP-binding cassette, Subfamily B, Member 6 protein that segregated with FP in the Cardiff family and was also present in both blood donors. Arg723Gln is listed in human variation databases with an allele frequency of approximately 1:1000. CONCLUSIONS: We describe a novel FP mutation that may affect 1:500 European blood donors and causes rapid loss of potassium from stored RBCs. This finding has implications for neonates and infants receiving large-volume RBC transfusions. Genomic screening of donors could be used to identify donors with this mutation and potentially improve the quality and safety of donor units.
T
he characteristics of stored red blood cells (RBCs) for transfusion depend on both the storage conditions and the physiology of the donated RBCs.1 Variability in the donated RBCs could be due to acquired changes related to diet or other factors or due to genetic variability among blood donors. Here, we report two cases in which the quality of a donor unit of RBCs was compromised by a genetically determined temperature-dependent potassium leak, causing hyperkalemia in the supernatant of the stored RBCs. ABBREVIATIONS: ABC = ATP-binding cassette; ABCB6 = ABC transporter Subfamily B Member 6; FP = familial pseudohyperkalemia; RT = room temperature; SNP = single-nucleotide polymorphism. From the 1Bristol Institute for Transfusion Sciences, NHS Blood & Transplant, and the 2School of Biochemistry, University of Bristol, Bristol, UK; 3Newcastle upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, UK; the 4Department of Haematology, University of Cambridge, NHS Blood & Transplant, & NIHR Biomedical Research Centre, Cambridge, UK; the 5Components Development Laboratory, NHS Blood & Transplant, Brentwood, UK; 6Imperial College Healthcare NHS Trust/NHS Blood & Transplant, and 7NHS Blood & Transplant Colindale Centre, London, UK; 8NHS Blood & Transplant Newcastle Centre, Newcastle upon Tyne, UK; and the 9School of Life and Medical Sciences, University College London, London, UK. Address correspondence to: Lesley J. Bruce, Bristol Institute for Transfusion Sciences, NHS Blood and Transplant, North Bristol Park, Filton, Bristol, BS34 7QH, UK; e-mail: [email protected]. The work was supported by the UK National Health Service R & D Directorate (JFF, RC, LJB). WMB was supported by a scholarship from the King Abdulaziz University, Faculty of Applied Medical Sciences, Jeddah, Saudi Arabia. Received for publication December 28, 2013; revision received April 7, 2014, and accepted May 13, 2014. doi: 10.1111/trf.12757 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
BAWAZIR ET AL.
Abnormal RBC membrane cation leakage is the hallmark of the hereditary stomatocytosis group of conditions. These conditions have been attributed to a variety of RBC membrane protein defects2-6 and are associated with a spectrum of clinical disorders ranging from severe hemolytic anemia to clinically benign.7 The mildest of these disorders are due to excessive potassium leakage that becomes apparent only when cells are stored at temperatures below 37°C and have been termed “familial pseudohyperkalemia” (FP).8 Some forms of FP have recently been shown to be associated with singlenucleotide substitutions affecting Amino Acid 375 of the ATP-binding cassette (ABC) transporter Subfamily B Member 6 (ABCB6).6 ABCB6 is a 94-kDa protein ubiquitously expressed in human tissues; it is a multispanning membrane protein that forms homodimers and is thought to be involved in porphyrin transport.9 It is highly expressed during erythroid differentiation and carries the high-incidence antigen Langereis (LAN) on the mature RBC membrane.9 ABCB6 is a newly described RBC membrane protein and its precise function in the RBC remains uncertain. However, the recent finding that mutations associated with FP may affect the ABCB6 protein structure6 demonstrates that this protein in abnormal form is linked with dysregulated cation transport in the cold. In this report we investigate two blood donors together with a recognized family with FP, previously reported as FP-Cardiff.10 We show that the two donors and the affected members of the FP-Cardiff family carry a recognized single-nucleotide polymorphism (SNP) in ABCB6 that had not previously been linked to a disease or phenotype. This relatively common SNP may have important implications for blood transfusion.
CASE REPORTS AND METHODS Case reports
of the returned RBC unit was found to have a potassium of 33.9 mmol/L. There was only minimal hemolysis, insufficient to account for the increased potassium. As a result the donor was recalled and further tests were undertaken. The donor was a 65-year-old male who had previously donated 60 units without any reported adverse effects on recipients. Further samples were obtained for genetic testing with the donor’s consent.
Case 2 A scenario identical to that of Case 1 occurred at a second hospital. A unit of RBCs from a regular donor from London accredited for neonatal use was used to prime a bypass machine for cardiac surgery on a 4-month-old infant. The unit was Day 5 postdonation and had not been irradiated. A blood gas measurement was taken from the bypass circuit before attaching the patient, and the potassium was 13.76 mmol/L. The unused part of the unit was tested and the supernatant was found to contain 41.4 mmol/L potassium. There was no evidence of excess hemolysis. The donor was a 54-year-old male who had previously donated 44 units without any recorded adverse events. Samples were obtained for genetic testing with the donor’s consent.
FP-Cardiff family The phenotype of the Cardiff family first reported with this type of FP (FP-Cardiff) has previously been described in detail.10 In brief, they were recognized by the finding of significant hyperkalemia after routine blood sampling. Laboratory studies found that the RBCs showed a greatly increased permeability to cations at refrigerator temperatures.10 Apart from pseudohyperkalemia, no clinical associations had been linked to this in vitro abnormality and in particular frank hemolysis was not present. In this study, samples were obtained for genetic testing from three affected and five unaffected members with the family’s consent.
Case 1 A unit of RBCs in optimal additive solution (SAG-M; quadruple top-top pack, Macopharma, Tourcoing, France) was donated by a regular donor from Cumbria who was accredited for neonatal use. The unit, which was Day 6 postdonation and had not been irradiated, was selected to be used for a neonate undergoing cardiac bypass that required priming of the extracorporeal circuit with RBCs before connection with the infant’s circulation. Units for this purpose are normally selected as being no more than 5 days postdonation11 to minimize the extracellular level of potassium. As part of routine practice the perfusionist preparing the circuit measured sodium and potassium in a blood gas analyzer before connecting the circuit with the infant. The supernatant potassium of the circuit was found to be unacceptably high and the unused part of the unit of RBCs was sent for further analysis. The supernatant 2
TRANSFUSION Volume **, ** **
Measurement of extracellular potassium concentration Lithium-heparin–anticoagulated blood samples from the two blood donors and two controls were stored at 20 or 4°C over 1 or 3 days and sampled periodically. A leukoreduced unit of RBCs in SAG-M from the London donor was stored at 4°C for 2 weeks and sampled periodically. Samples removed from this unit were either left at 4°C (untreated) or incubated for 2 hours in a 37°C water bath, with the addition of Li-pyruvate and inosine (both at 15 mmol/L) to encourage reactivation of the NaKATPase and restoration of the cation gradients (rejuvenated). Extracellular potassium concentration was measured by flame photometry (BWB Technologies, Spectronic Camspec Ltd, Leeds, UK).
FAMILIAL PSEUDOHYPERKALEMIA IN BLOOD DONORS
I.1 (R723Q)
pean descent), and the UK10K whole-genome sequencing project (3621 individuals from the United Kingdom).
I .2 (R723)
Statistical analysis II.1 (R723)
II.2 (R723)
II.3 (R723Q) III.1 (R723Q)
II.4 (R723Q) III.2 (R723)
III.3 (n.t.)
III.4 (n.t.)
IV.1 (n.t.) Fig. 1. Family tree for the FP-Cardiff family. The FP-Cardiff phenotype was named after the first patient found with this phenotype (patient Cardiff, I-110). The FP-Cardiff phenotype has been reported in only one family from Cardiff but population density data on the SNP database suggest that this substitution may be more common. We report here this same substitution in two unrelated blood donors. The family tree shows the individuals with FP (solid symbols) and the unaffected individuals (open symbols). The heterozygous mutation (Arg723Gln) segregates with the condition. n.t. = not tested.
Reference data on potassium levels in units of RBCs in SAG-M stored according to NHS Blood and Transplant procedures were obtained by measuring supernatant potassium using a chemistry analyzer (Vitros DT 60 II, Ortho Clinical Diagnostics, High Wycombe, Bucks, UK). The data were collected between 2003 and 2006 by the Component Development Laboratory, NHSBT, as part of storage studies on RBCs during validation of systems for processing whole blood or RBCs that were all approved for routine use. These studies assessed RBCs at weekly intervals throughout storage, that is, Days 0 or 1, 7, 14, 21, and 35. Further, since levels on Day 5 of storage are of particular interest in relation to neonatal transfusion, reference data on units of RBCs stored in CPD plasma were also measured on Day 5 of storage.
DNA analysis Genomic DNA was isolated from blood samples. For the proband of the FP-Cardiff family (I-1, Fig. 1) all coding regions of human ABCB6 were amplified by polymerase chain reaction and the DNA was sequenced as described previously.2 DNA sequencing was used to analyze exon 16 of ABCB6 from all available family members and from the two blood donors. Allele frequencies were obtained from the database of SNPs (dbSNP) and dbSNP rel 137. Frequency data were available from four studies, 1000 Genomes (1092 individuals), the NHLBI Exome Sequencing Project (2203 African American and 4300 European Americans), the ClinSeq project (514 individuals of Euro-
Confidence intervals (CIs) of proportions and standard deviations (SDs) were calculated using statistics software (GraphPad, GraphPad Software, Inc., La Jolla, CA). The mean, median, and percentile analysis was calculated using computer software (GraphPad Prism 5; or Microsoft Excel, Microsoft, London, UK).
RESULTS FP-Cardiff is associated with Arg723Gln substitution in ABCB6 Pseudohyperkalemia was confirmed in both donors; their blood plasma potassium after storage was only slightly increased at 20°C but increased much more rapidly than the control when stored at 4°C. Plasma potassium concentration in Li-heparin–anticoagulated samples from the London donor was measured over 3 days (Fig. 2A). Plasma potassium concentration in Li-heparin–anticoagulated samples from the Cumbrian donor was measured immediately and after an overnight hold at room temperature (RT) and at 4°C alongside a control sample from a second individual. The plasma potassium was 4.7 mmol/L immediately after sampling, 5.0 mmol/L after 18 hours at RT, and 17.4 mmol/L after 18 hours at 4°C. The control sample showed respective values of 5.1, 3.7, and 8.6 mmol/L. These findings were very similar in pattern and degree to the potassium permeability previously described for FP-Cardiff.10 Supernatant potassium concentration in a unit of RBCs (leukoreduced and in SAG-M) from the London donor increased rapidly over 14 days’ storage at 4°C (Fig. 2B). By Day 1, at 17.5 mmol/L it was above the median potassium concentration found on Day 5 in standard RBCs (Fig. 3A); by Day 4 it was 43.1 mmol/L (Fig. 2B), a potassium concentration more usually found on Day 35 of storage (Fig. 3A). Rejuvenating the RBCs, by incubation at 37°C for 2 hours in the presence of inosine and pyruvate, reduced the supernatant potassium concentration by approximately 10 mmol/L, caused by reuptake of a fraction of the leaked potassium. This reduction of approximately 10 mmol/L was consistent irrespective of storage time, and on Day 4 warming the RBCs reduced the supernatant potassium from 43.1 to 33.1 mmol/L (Fig. 2B), a potassium concentration more usually found on Day 21 of storage (Fig. 3A). Potassium levels in standard units of RBCs prepared for transfusion were measured and the increase during storage plotted (Fig. 3A). From the data, the supernatant of Day 5 stored RBCs in SAG-M would be estimated to contain a mean of approximately 11 to 12 mmol/L Volume **, ** **
TRANSFUSION
3
BAWAZIR ET AL.
A
B
22
65
20
60
K in supernatant (mmol/L)
18
K in plasma (mmol/L)
Population frequency of the Arg723Gln SNP
70
16 14 12 10 8
55 50 45 40 35 30 25 20 15
6
10
4
5 0
10 20 30 40 50 60 70 80
Time (hr)
0
2
4
6
8
10
12
14
Time (days)
Fig. 2. Cation permeability of RBCs from the London blood donor. (A) A heparinanticoagulated blood sample from the London donor was incubated over 3 days at
The Arg723Gln substitution in ABCB6 is listed on dbSNP with identifier rs148211042 but is not linked to a disease or phenotype. The 1000 Genomes, UK10K, NHLBI-ESP, and the ClinSeq projects all provide frequency data for rs14811042. Overall the data show that this SNP is present at an allele frequency of approximately 1:1000 in European populations (22 in 17,875 alleles observed [Table 1]) such that the heterozygous state will be present in approximately 1 in 500 European blood donors (95% CI, 1 in 270 to 1 in 625).
DISCUSSION
RBC membrane cation leaks that are temperature dependent have been RT (open symbols) or 4°C (closed symbols) and the plasma potassium concentration described in a variety of disorders was measured by flame photometry. (▽) Control; (▼) control 4°C; (○) London donor; ranging from moderately severe hemo(●) London donor 4°C. (B) A unit of leukoreduced RBCs in SAG-M from the London lytic states to asymptomatic FP.7 donor was stored for 2 weeks at 4°C. Samples were taken on Days 1, 4, 7, and 14 and The disorders are dominantly inherited either left at 4°C (untreated) or incubated at 37°C for 2 hours in the presence of with complete penetrance. Clinically inosine and pyruvate (rejuvenated). Extracellular potassium concentration was measignificant forms include Southeast sured by flame photometry. (○) London donor untreated; (●) London donor rejuveAsian ovalocytosis,12 dehydrated and nated. Error bars show the SD from the mean (n = 3). overhydrated stomatocytosis, and cryohydrocytosis.7,13 Cases with a clinical phenotype are generally very rare in European popupotassium (Fig. 3A). This is consistent with that measured lations and will often render an affected individual in units of RBCs stored in CPD plasma on Day 5 where unsuitable for blood donation due to anemia or other median supernatant potassium levels were 11.94 mmol/L problems. FP, which has been described in a number of (range, 6.00-21.60 mmol/L; n = 61). The data from Day 7 kindreds, is clinically benign and probably underdiagon 177 units shows a skewed distribution with a median nosed. Asymptomatic individuals with FP may therefore value of 14.0 mmol/L and 2.5th and 97.5th percentiles of become blood donors. Several different forms of FP have 6.1 and 31.8 mmol/L, respectively (Fig. 3B). The units of been described with cation leaks that vary in their temRBCs in the two case reports described here were on Day 5 perature dependence (illustrated in Fig. 4).7 For instance or Day 6 of storage and the supernatant potassium in both cases (33.9 and 41.4 mmol/L, respectively) was well above FP-Edinburgh is a dehydrated stomatocytosis or the expected level for RBCs at that time in storage and xerocytosis type, caused by a mutation in PIEZO114 that above the 97.5th percentile even for the data set on Day 7 has a minor cation leak at RT but does not leak at 4°C (Fig. 3B). (Fig. 4A).7 The FP-Chiswick type, which includes the famiGenomic DNA was analyzed from the FP-Cardiff lies FP-Falkirk, FP-East London (Bow), and FP-Lille, family and from the two blood donors. The NM_005689.2: caused by the substitutions Arg375Trp and Arg375Gln in c.2168G>A mutation leading to the substitution ABCB6 reported by Andolfo and colleagues,6 leaks cations p.Arg723Gln was found in the proband of the FP-Cardiff both at 20°C and at 4°C (Fig. 4B).7,15 However, neither family (I.1; Fig. 1). DNA analysis of an additional seven Arg375 mutation is listed in the dbSNP, and these mutamembers of the FP-Cardiff family showed that the tions are probably very rare. The FP-Cardiff type described NM_005689.2:c.2168G>A mutation segregates with FP in here has RBC cation permeability that is normal at 37°C, the Cardiff family. The three affected FP-Cardiff family only mildly abnormal at 20°C, but at 4°C it is six times that members were heterozygous for the substitution and the of control RBCs (Fig. 4C).7,10 five unaffected members expressed only Arg723 (Fig. 1). In this study, we describe a novel FP mutation in the The same mutation and p.Arg723Gln substitution was also ABCB6 gene that cosegregates with FP in a family from found in both blood donors. Cardiff (FP-Cardiff) and results in the heterozygous 4
TRANSFUSION Volume **, ** **
FAMILIAL PSEUDOHYPERKALEMIA IN BLOOD DONORS
70 60 50 40 30 20 10
40 30 20
Mean (15.5)
42 ay
D
ay
35
28 D
ay
21
2.5% Median (6.1) (14.0)
D
ay
14 D
ay D
ay
1
50
D
D
Number of un its
B
7
0
ay
Supern natant Potassiu um (mmol/L)
A
97.5% (31.8)
Case 1 Day 6 33mmol/L Case 2 Day 5 41mmol/L
10
4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42
0
Supernatant potassium (mmol/L) Fig. 3. Supernatant potassium concentration of stored donor RBC units. (A) Supernatant potassium concentration of RBC units over 42 days. The supernatant potassium of RBCs is routinely monitored for quality by the Components Development Laboratory (CDL) NHSBT. Data presented were measured over 42 days of refrigerated storage in randomly selected units between 2003 and 2006. The data for each time point are represented by the median (solid bar), the 25th and 75th percentiles of the median (box), and the 2.5th and 97.5th percentiles of the median (error bars). Day 1, n = 538; Day 7, n = 177; Day 14, n = 137; Day 21, n = 69; Day 28, n = 29; Day 35, n = 173; Day 42, n = 119. (B) Supernatant potassium concentration of RBC units on Day 7 of storage. The graph shows the distribution of the potassium concentration in RBC units on Day 7 of refrigerated storage as collected by CDL, NHSBT. Data represent randomly selected units collected between 2003 and 2006 (n = 177). The distribution of the data is not normal but is right-skewed. Both the median and median percentile data and the mean data are shown above the bar chart. The potassium concentration of each of the FP-Cardiff type donor units reported in this article was measured on Day 5 or Day 6 so are not directly comparable to the Day 7 CDL data; however, both these units would fall within the right-hand tail of the distribution at positions indicated by arrows.
substitution (Arg723Gln) in ABCB6 protein. The same SNP was found in the two blood donors with unusually high potassium concentrations (33.9 and 41.4 mmol/L) in recently stored donations. Pseudohyperkalemia was confirmed in the donors; their blood plasma potassium after storage was only slightly raised at 20°C but increased much more rapidly than the control when stored at 4°C identical to the changes previously recorded for FP-Cardiff.10 Unlike other described mutations in ABCB66 and PIEZO114 causing FP, this newly described mutation is relatively common with a frequency of this dominant trait of approximately 1 in 500 in persons of European origin. Potassium in the supernatant of stored cells increases during storage (Fig. 3A). For the majority of transfusions this is probably of limited significance because the total amount of potassium transfused is relatively small compared to the total blood volume of the recipient. However, there are certain clinical scenarios where transfusion of high-potassium donor blood may be more of a concern. In large-volume neonatal and pediatric transfusion, high levels of supernatant potassium can have serious or fatal consequences.16-20 The level of transfused supernatant potassium that could be potentially dangerous to a recipient would depend on the rate of transfusion, the relative volume of transfusion compared to the neonatal or infant blood volume, and other contributory factors such as the clinical status of the recipient (recently reviewed by Lee et al.20). For example, a study by Brown and colleagues21 suggested in a simulated model that low cardiac output states can be associated with hyperkalemia if the potassium concentration in the supernatant of the rapidly transfused blood exceeds a level as low as 10 mmol/L. It is common for blood services to restrict the shelf life of RBCs for large-volume neonatal transfusion to reduce the risk of hyperkalemia, and some hospitals may wash RBCs to reduce potassium levels before transfusion. Most hospital departments in the United Kingdom using blood for neonatal bypass prime measure the potassium of the prime fluid before use, although perfusion practice is variable.22 However, blood for large-volume transfusion in neonates in other circumstances, although stored for a short period only, is not routinely tested for supernatant potassium. The RBC unit from the London donor described here contained 17.5 mmol/L potassium on Day 1 of storage, and the potassium level rapidly increased to 43.1 mmol/L by Day 4 of storage, a level commonly found on Day 35 of storage in standard RBC units. After rejuvenation of the RBCs we found some reduction in supernatant potassium in the FP donor unit. For example, after 4 days of storage at 4°C rejuvenation reduced the supernatant potassium from 43.1 to 33.1 mmol/L, but this is a level that would still be considered potentially dangerous for a large-volume transfusion to a child. RBC units with unexpectedly high potassium Volume **, ** **
TRANSFUSION
5
BAWAZIR ET AL.
TABLE 1. Database analysis of allele frequency* European (allele frequency) 9:8,591 1:757 10:7,500 2:1,027 22:17,875 (0.1%)
Database NHLBI Exome Sequencing Project 1000 Genomes project UK10K project CSAgilent ClinSeq project Total
African American (allele frequency) 2:4404
0.14
C
0.12 0.10 0.08 0.06 0.04 0.02 0.00 40
30
20
10
Temperature (°C)
0
Ouabain+bumetanide K influx (mmol/L cells/h)
B Ouabain+bumetanide K influx (mmol/L cells/h)
A
Ouabain+bumetanide K influx (mmmol/L cells/h)
* Frequency data were available from four databases, 1000 Genomes (1092 individuals), the NHLBI Exome Sequencing Project (2203 African American and 4300 European Americans), the ClinSeq project (514 individuals of European descent), and the UK10K wholegenome sequencing project (3621 individuals from the UK). Overall the data suggest an allele frequency of 1:1000. As the heterozygous p.R723Q (c.g2168a) mutation is dominantly inherited with complete penetrance this suggests that this mutation may be present in 1:500 individuals of European descent.
0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 40
30
20
10
Temperature (°C)
0
0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 40
30
20
10
0
Temperature (°C)
Fig. 4. Illustration showing the temperature dependence of the cation leak in different FP phenotypes. These data are shown only as an illustration and are redrawn from Stewart.7 The method measured potassium influx using 86Rb as a tracer in the presence of 0.1 mmol/L ouabain and bumetanide as described.10 The data illustrate the difference in the temperature dependence of the cation leak between the three FP phenotypes. FP-Cardiff RBCs are unique in that their cation permeability is almost identical to control RBCs at 37°C but is six times normal at 0°C. (A) Temperature dependence of potassium flux in FP-Edinburgh RBCs. (○) Control; (●) FP-Edinburgh. (B) Temperature dependence of potassium flux in FP-Chiswick RBCs. (○) Control; (●) FP-Chiswick. (C) Temperature dependence of potassium flux in FP-Cardiff RBCs. (○) Control; (●) FP-Cardiff.
from donors such as we have described could potentially cause significant hyperkalemia in these patients.20,23,24 There is increasing awareness that inherited changes in RBCs or platelets (PLTs) may affect survival and function of transfused cells.25 Here, we have described a significant RBC abnormality due to a relatively common polymorphism that causes no known clinical problems in the donor. Potassium leaks may be detected in affected individuals when unexpected hyperkalemia occurs in blood samples kept overnight at RT before testing. However, when hyperkalemia is detected in such patients, the plasma potassium is typically repeated urgently on a fresh sample and found to be normal. The original abnormal potassium is then dismissed as a laboratory error or 6
TRANSFUSION Volume **, ** **
attributed to lysis. Therefore, these donors can go unrecognized both clinically as patients and as blood donors. Standard RBC units stored for up to 35 days in the United Kingdom, or 42 days elsewhere, may have supernatant potassium levels of 50 mmol/L but total potassium loads of only approximately 5 mmol/unit due to the relatively small volume of the supernatant. It is to be expected that there will be variation in the level of potassium in the supernatant of RBC units, determined by factors related both to the donor and to changes in the RBCs due to processing and storage; however, published data on this distribution in large data sets are lacking. We have shown that in standard RBC units during storage there is a wide range of values of supernatant potassium. Supernatant
FAMILIAL PSEUDOHYPERKALEMIA IN BLOOD DONORS
potassium concentration in units stored for 7 days mainly fall within the 2.5th to 97.5th percentile range (6.131.8 mmol/L) but there is a tail end of units that have levels above the 97.5 percentile from the median (Fig. 3B). Our data demonstrate that donors with the FP mutation may contribute to the latter data on Day 5 or 6 of storage. However, potassium levels in RBC units from these donors will not be easily distinguishable from those of normal donors during the later weeks of normal RBC storage when potassium levels reach a plateau (Fig. 3A; Fig. S1, available as supporting information in the online version of this paper). The clinical significance of the FP mutation for RBC transfusion is likely to be greatest for large-volume transfusions to neonates, where units are deliberately selected to be no more than 5 days postdonation, as units from donors with the mutation are likely to have potassium levels that are well above average for the unit age. However, it is possible that there could be additional effects, such as altered RBC functionality or decreased stability of the RBC membrane causing increased vesiculation, that could occur with increasing lengths of storage in RBCs units from donors with FP-Cardiff and this is an area of ongoing investigation. Having identified the FP-Cardiff mutation and highlighted the potential clinical significance, it will be important to determine the genotype and frequency of other FP causing mutations. The development of a method that could reliably identify those individuals bearing the common FP mutations in the donor population could be considered alongside other genomic donor screening tests for blood groups, HLA types, and other genetic variations as an effective way of improving the quality of the RBC supply and the provision of highly selected units for specific patient groups in the future. In conclusion, we report here a novel mutation that may affect 1:500 clinically asymptomatic European blood donors and that leads to an increased rate of potassium leakage from RBCs when stored at 4°C. Recognition and testing for genetic variants affecting RBC or PLT transfusion physiology such as described here could help improve the quality of the blood supply. We are currently undertaking a series of further studies to fully elucidate the implications of this finding for RBC transfusion and to inform policy on how to manage donors who may be identified with the FP mutation.
research design; and all authors contributed to the manuscript preparation.
WEB-BASED RESOURCES (ACCESSED IN AUGUST 2013) dbSNP rel 137 (http://bioq.saclab.net/query/submit.php ?db=bioq_dbsnp_human_137) dbSNP (http://www.ncbi.nlm.nih.gov/snp). 1000 Genomes (http://browser.1000genomes.org). NHLBI Exome Sequencing Project (http://evs.gs .washington.edu/). ClinSeq project (http://www.genome.gov/20519355). UK10K whole-genome sequencing project (http:// www.uk10k.org/). CONFLICT OF INTEREST NHSBT has filed a patent “Method of detecting cation leaks” Patent No GB1219226.6. Apart from the patent, the authors have disclosed no conflicts of interest.
REFERENCES 1. Hess JR. Red cell changes during storage. Transfus Apher Sci 2010;43:51-9. 2. Bruce LJ, Robinson HC, Guizouarn H, et al. Monovalent cation leaks in human red cells caused by single aminoacid substitutions in the transport domain of the band 3 chloride-bicarbonate exchanger, AE1. Nat Genet 2005;37: 1258-63. 3. Bruce LJ, Guizouarn H, Burton NM, et al. The monovalent cation leak in overhydrated stomatocytic red blood cells results from amino acid substitutions in the Rh-associated glycoprotein. Blood 2009;113:1350-7. 4. Flatt JF, Guizouarn H, Burton NM, et al. Stomatin-deficient cryohydrocytosis results from mutations in SLC2A1: a
5.
6.
7.
8. ACKNOWLEDGMENTS We thank J. Hosken for DNA sequencing. We thank the donors, patients, and their families for their cooperation. WMB performed DNA and cation analysis; JFF, MW, and RAC performed cation analysis; AR performed population genetics analyses; CEC, LP, and HVN carried out provision of donor samples; GWS carried out provision of patient samples; LJB and JPW performed
9.
10.
novel form of GLUT1 deficiency syndrome. Blood 2011; 118:5267-77. Zarychanski R, Schulz VP, Houston BL, et al. Mutations in the mechanotransduction protein PIEZO1 are associated with hereditary xerocytosis. Blood 2012;120:1908-15. Andolfo I, Alper SL, Delaunay J, et al. Missense mutations in the ABCB6 transporter cause dominant familial pseudohyperkalemia. Am J Hematol 2013;88:66-72. Stewart GW. Hemolytic disease due to membrane ion channel disorders. Curr Opin Hematol 2004;11: 244-50. Stewart GW, Corrall RJ, Fyffe JA, et al. Familial pseudohyperkalaemia. A new syndrome. Lancet 1979;2: 175-7. Helias V, Saison C, Ballif BA, et al. ABCB6 is dispensable for erythropoiesis and specifies the new blood group system Langereis. Nat Genet 2012;44:170-3. Gore DM, Chetty MC, Fisher J, et al. Familial pseudohyperkalaemia Cardiff: a mild version of cryohydrocytosis. Br J Haematol 2002;117:212-4. Volume **, ** **
TRANSFUSION
7
BAWAZIR ET AL.
11. National Blood Service. Guidelines for the blood transfusion services in the UK. 8th ed. London, UK: The Stationery Office; 2013. [cited 2013 Aug]. Available from: http:// www.transfusionguidelines.org.uk/red-book 12. Guizouarn H, Borgese F, Gabillat N, et al. South-east Asian ovalocytosis and the cryohydrocytosis form of hereditary stomatocytosis show virtually indistinguishable cation permeability defects. Br J Haematol 2011;152:655-64. 13. Delaunay J. The hereditary stomatocytoses: genetic disorders of the red cell membrane permeability to monovalent cations. Semin Hematol 2004;41:165-72. 14. Andolfo I, Alper SL, De Franceschi L, et al. Multiple clinical forms of dehydrated hereditary stomatocytosis arise from
23. Vraets A, Lin Y, Callum JL. Transfusion-associated hyperkalemia. Transfus Med Rev 2011;25:184-96. 24. Strauss RG. RBC storage and avoiding hyperkalemia from transfusions to neonates & infants. Transfusion 2010;50: 1862-5. 25. Kleinman S, Busch MP, Murphy EL, et al. The National Heart, Lung, and Blood Institute Recipient Epidemiology and Donor Evaluation Study (REDS-III). The National Heart, Lung, and Blood Institute Recipient Epidemiology and Donor Evaluation Study (REDS-III): a research program striving to improve blood donor and transfusion recipient outcomes. Transfusion 2013;54(3 Pt 2):942-55.
mutations in PIEZO1. Blood 2013;121:3925-35. 15. Carella M, d’Adamo AP, Grootenboer-Mignot S, et al. A second locus mapping to 2q35-36 for familial pseudohyperkalaemia. Eur J Hum Genet 2004;12:1073-6. 16. Hall TL, Barnes A, Miller JR, et al. Neonatal mortality following transfusion of red cells with high plasma potassium levels. Transfusion 1993;33:606-9. 17. Chen CH, Hong CL, Kau YC, et al. Fatal hyperkalemia during rapid and massive blood transfusion in a child undergoing hip surgery—a case report. Acta Anaesthesiol Sin 1999;37:163-6. 18. Baz EM, Kanazi GE, Mahfouz RA, et al. An unusual case of hyperkalaemia-induced cardiac arrest in a paediatric patient during transfusion of a “fresh” 6-day-old blood unit. Transfus Med 2002;12:383-6. 19. Smith HM, Farrow SJ, Ackerman JD, et al. Cardiac arrests associated with hyperkalemia during red blood cell transfusion: a case series. Anesth Analg 2008;106:1062-9. 20. Lee AC, Reduque LL, Luban NL, et al. Transfusionassociated hyperkalemic cardiac arrest in pediatric patients receiving massive transfusion. Transfusion 2014; 54:244-54. 21. Brown KA, Bissonnette B, McIntyre B. Hyperkalaemia during rapid blood transfusion and hypovolaemic cardiac arrest in children. Can J Anaesth 1990;37:747-54. 22. Harvey B, Shann KG, Fitzgerald D, et al.; American Society of ExtraCorporeal Technology’s International Consortium for Evidence-Based Perfusion and Pediatric Perfusion Committee. International pediatric perfusion practice: 2011 survey results. J Extra Corpor Technol 2012;44: 186-93.
8
TRANSFUSION Volume **, ** **
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s Web site: Fig. S1. Distribution of supernatant potassium concentration of stored donor red cell units. The graphs show the distribution of the potassium concentration in red cell units at weekly intervals of refrigerated storage as collected by CDL, NHSBT. Data represents randomly selected units collected between 2003-2006. The distribution of the data varies with the week of storage. The median and median percentile data are shown above the bar chart. The potassium concentration of each of the FP-Cardiff type donor units reported in this paper was measured on Day 5 or Day 6 and are shown on the Day 7 graphs; both these units fall within the righthand tail of the distribution at positions indicated by arrows. A) Distribution of the supernatant potassium concentration of red cell units on Day 0/Day 1 of storage. B) Distribution of the supernatant potassium concentration of red cell units on Day 7 of storage. C) Distribution of the supernatant potassium concentration of red cell units on Day 14 of storage. D) Distribution of the supernatant potassium concentration of red cell units on Day 21 of storage. E) Distribution of the supernatant potassium concentration of red cell units on Day 35 of storage. F) Distribution of the supernatant potassium concentration of red cell units on Day 42 of storage.
BACKGROUND: Familial pseudohyperkalemia (FP) is a dominantly inherited condition in which red blood cells (RBCs) have an increased cold-induced permeability to monovalent cations. Potassium leaks into the supernatant of all stored blood with time, but FP RBCs leak potassium more rapidly. We investigated two unrelated blood donors whose RBC donations demonstrated unexpectedly high potassium after 5 and 6 days’ storage. We matched the observed pattern of RBC cation leak to a previously recognized family with FP (FP-Cardiff) and investigated the likely cause with targeted DNA analysis. STUDY DESIGN AND METHODS: Cation leakage from the donor RBCs and from standard donor units was measured. DNA analysis of donors and family members with FP-Cardiff was performed. Allele frequencies were obtained from human variation databases. RESULTS: Both implicated donors were found to have increased cold-induced potassium leak identical in pattern to affected members of the family with FP-Cardiff. We found a heterozygous substitution Arg723Gln in the ATP-binding cassette, Subfamily B, Member 6 protein that segregated with FP in the Cardiff family and was also present in both blood donors. Arg723Gln is listed in human variation databases with an allele frequency of approximately 1:1000. CONCLUSIONS: We describe a novel FP mutation that may affect 1:500 European blood donors and causes rapid loss of potassium from stored RBCs. This finding has implications for neonates and infants receiving large-volume RBC transfusions. Genomic screening of donors could be used to identify donors with this mutation and potentially improve the quality and safety of donor units.
T
he characteristics of stored red blood cells (RBCs) for transfusion depend on both the storage conditions and the physiology of the donated RBCs.1 Variability in the donated RBCs could be due to acquired changes related to diet or other factors or due to genetic variability among blood donors. Here, we report two cases in which the quality of a donor unit of RBCs was compromised by a genetically determined temperature-dependent potassium leak, causing hyperkalemia in the supernatant of the stored RBCs. ABBREVIATIONS: ABC = ATP-binding cassette; ABCB6 = ABC transporter Subfamily B Member 6; FP = familial pseudohyperkalemia; RT = room temperature; SNP = single-nucleotide polymorphism. From the 1Bristol Institute for Transfusion Sciences, NHS Blood & Transplant, and the 2School of Biochemistry, University of Bristol, Bristol, UK; 3Newcastle upon Tyne Hospitals NHS Foundation Trust, Newcastle upon Tyne, UK; the 4Department of Haematology, University of Cambridge, NHS Blood & Transplant, & NIHR Biomedical Research Centre, Cambridge, UK; the 5Components Development Laboratory, NHS Blood & Transplant, Brentwood, UK; 6Imperial College Healthcare NHS Trust/NHS Blood & Transplant, and 7NHS Blood & Transplant Colindale Centre, London, UK; 8NHS Blood & Transplant Newcastle Centre, Newcastle upon Tyne, UK; and the 9School of Life and Medical Sciences, University College London, London, UK. Address correspondence to: Lesley J. Bruce, Bristol Institute for Transfusion Sciences, NHS Blood and Transplant, North Bristol Park, Filton, Bristol, BS34 7QH, UK; e-mail: [email protected]. The work was supported by the UK National Health Service R & D Directorate (JFF, RC, LJB). WMB was supported by a scholarship from the King Abdulaziz University, Faculty of Applied Medical Sciences, Jeddah, Saudi Arabia. Received for publication December 28, 2013; revision received April 7, 2014, and accepted May 13, 2014. doi: 10.1111/trf.12757 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
BAWAZIR ET AL.
Abnormal RBC membrane cation leakage is the hallmark of the hereditary stomatocytosis group of conditions. These conditions have been attributed to a variety of RBC membrane protein defects2-6 and are associated with a spectrum of clinical disorders ranging from severe hemolytic anemia to clinically benign.7 The mildest of these disorders are due to excessive potassium leakage that becomes apparent only when cells are stored at temperatures below 37°C and have been termed “familial pseudohyperkalemia” (FP).8 Some forms of FP have recently been shown to be associated with singlenucleotide substitutions affecting Amino Acid 375 of the ATP-binding cassette (ABC) transporter Subfamily B Member 6 (ABCB6).6 ABCB6 is a 94-kDa protein ubiquitously expressed in human tissues; it is a multispanning membrane protein that forms homodimers and is thought to be involved in porphyrin transport.9 It is highly expressed during erythroid differentiation and carries the high-incidence antigen Langereis (LAN) on the mature RBC membrane.9 ABCB6 is a newly described RBC membrane protein and its precise function in the RBC remains uncertain. However, the recent finding that mutations associated with FP may affect the ABCB6 protein structure6 demonstrates that this protein in abnormal form is linked with dysregulated cation transport in the cold. In this report we investigate two blood donors together with a recognized family with FP, previously reported as FP-Cardiff.10 We show that the two donors and the affected members of the FP-Cardiff family carry a recognized single-nucleotide polymorphism (SNP) in ABCB6 that had not previously been linked to a disease or phenotype. This relatively common SNP may have important implications for blood transfusion.
CASE REPORTS AND METHODS Case reports
of the returned RBC unit was found to have a potassium of 33.9 mmol/L. There was only minimal hemolysis, insufficient to account for the increased potassium. As a result the donor was recalled and further tests were undertaken. The donor was a 65-year-old male who had previously donated 60 units without any reported adverse effects on recipients. Further samples were obtained for genetic testing with the donor’s consent.
Case 2 A scenario identical to that of Case 1 occurred at a second hospital. A unit of RBCs from a regular donor from London accredited for neonatal use was used to prime a bypass machine for cardiac surgery on a 4-month-old infant. The unit was Day 5 postdonation and had not been irradiated. A blood gas measurement was taken from the bypass circuit before attaching the patient, and the potassium was 13.76 mmol/L. The unused part of the unit was tested and the supernatant was found to contain 41.4 mmol/L potassium. There was no evidence of excess hemolysis. The donor was a 54-year-old male who had previously donated 44 units without any recorded adverse events. Samples were obtained for genetic testing with the donor’s consent.
FP-Cardiff family The phenotype of the Cardiff family first reported with this type of FP (FP-Cardiff) has previously been described in detail.10 In brief, they were recognized by the finding of significant hyperkalemia after routine blood sampling. Laboratory studies found that the RBCs showed a greatly increased permeability to cations at refrigerator temperatures.10 Apart from pseudohyperkalemia, no clinical associations had been linked to this in vitro abnormality and in particular frank hemolysis was not present. In this study, samples were obtained for genetic testing from three affected and five unaffected members with the family’s consent.
Case 1 A unit of RBCs in optimal additive solution (SAG-M; quadruple top-top pack, Macopharma, Tourcoing, France) was donated by a regular donor from Cumbria who was accredited for neonatal use. The unit, which was Day 6 postdonation and had not been irradiated, was selected to be used for a neonate undergoing cardiac bypass that required priming of the extracorporeal circuit with RBCs before connection with the infant’s circulation. Units for this purpose are normally selected as being no more than 5 days postdonation11 to minimize the extracellular level of potassium. As part of routine practice the perfusionist preparing the circuit measured sodium and potassium in a blood gas analyzer before connecting the circuit with the infant. The supernatant potassium of the circuit was found to be unacceptably high and the unused part of the unit of RBCs was sent for further analysis. The supernatant 2
TRANSFUSION Volume **, ** **
Measurement of extracellular potassium concentration Lithium-heparin–anticoagulated blood samples from the two blood donors and two controls were stored at 20 or 4°C over 1 or 3 days and sampled periodically. A leukoreduced unit of RBCs in SAG-M from the London donor was stored at 4°C for 2 weeks and sampled periodically. Samples removed from this unit were either left at 4°C (untreated) or incubated for 2 hours in a 37°C water bath, with the addition of Li-pyruvate and inosine (both at 15 mmol/L) to encourage reactivation of the NaKATPase and restoration of the cation gradients (rejuvenated). Extracellular potassium concentration was measured by flame photometry (BWB Technologies, Spectronic Camspec Ltd, Leeds, UK).
FAMILIAL PSEUDOHYPERKALEMIA IN BLOOD DONORS
I.1 (R723Q)
pean descent), and the UK10K whole-genome sequencing project (3621 individuals from the United Kingdom).
I .2 (R723)
Statistical analysis II.1 (R723)
II.2 (R723)
II.3 (R723Q) III.1 (R723Q)
II.4 (R723Q) III.2 (R723)
III.3 (n.t.)
III.4 (n.t.)
IV.1 (n.t.) Fig. 1. Family tree for the FP-Cardiff family. The FP-Cardiff phenotype was named after the first patient found with this phenotype (patient Cardiff, I-110). The FP-Cardiff phenotype has been reported in only one family from Cardiff but population density data on the SNP database suggest that this substitution may be more common. We report here this same substitution in two unrelated blood donors. The family tree shows the individuals with FP (solid symbols) and the unaffected individuals (open symbols). The heterozygous mutation (Arg723Gln) segregates with the condition. n.t. = not tested.
Reference data on potassium levels in units of RBCs in SAG-M stored according to NHS Blood and Transplant procedures were obtained by measuring supernatant potassium using a chemistry analyzer (Vitros DT 60 II, Ortho Clinical Diagnostics, High Wycombe, Bucks, UK). The data were collected between 2003 and 2006 by the Component Development Laboratory, NHSBT, as part of storage studies on RBCs during validation of systems for processing whole blood or RBCs that were all approved for routine use. These studies assessed RBCs at weekly intervals throughout storage, that is, Days 0 or 1, 7, 14, 21, and 35. Further, since levels on Day 5 of storage are of particular interest in relation to neonatal transfusion, reference data on units of RBCs stored in CPD plasma were also measured on Day 5 of storage.
DNA analysis Genomic DNA was isolated from blood samples. For the proband of the FP-Cardiff family (I-1, Fig. 1) all coding regions of human ABCB6 were amplified by polymerase chain reaction and the DNA was sequenced as described previously.2 DNA sequencing was used to analyze exon 16 of ABCB6 from all available family members and from the two blood donors. Allele frequencies were obtained from the database of SNPs (dbSNP) and dbSNP rel 137. Frequency data were available from four studies, 1000 Genomes (1092 individuals), the NHLBI Exome Sequencing Project (2203 African American and 4300 European Americans), the ClinSeq project (514 individuals of Euro-
Confidence intervals (CIs) of proportions and standard deviations (SDs) were calculated using statistics software (GraphPad, GraphPad Software, Inc., La Jolla, CA). The mean, median, and percentile analysis was calculated using computer software (GraphPad Prism 5; or Microsoft Excel, Microsoft, London, UK).
RESULTS FP-Cardiff is associated with Arg723Gln substitution in ABCB6 Pseudohyperkalemia was confirmed in both donors; their blood plasma potassium after storage was only slightly increased at 20°C but increased much more rapidly than the control when stored at 4°C. Plasma potassium concentration in Li-heparin–anticoagulated samples from the London donor was measured over 3 days (Fig. 2A). Plasma potassium concentration in Li-heparin–anticoagulated samples from the Cumbrian donor was measured immediately and after an overnight hold at room temperature (RT) and at 4°C alongside a control sample from a second individual. The plasma potassium was 4.7 mmol/L immediately after sampling, 5.0 mmol/L after 18 hours at RT, and 17.4 mmol/L after 18 hours at 4°C. The control sample showed respective values of 5.1, 3.7, and 8.6 mmol/L. These findings were very similar in pattern and degree to the potassium permeability previously described for FP-Cardiff.10 Supernatant potassium concentration in a unit of RBCs (leukoreduced and in SAG-M) from the London donor increased rapidly over 14 days’ storage at 4°C (Fig. 2B). By Day 1, at 17.5 mmol/L it was above the median potassium concentration found on Day 5 in standard RBCs (Fig. 3A); by Day 4 it was 43.1 mmol/L (Fig. 2B), a potassium concentration more usually found on Day 35 of storage (Fig. 3A). Rejuvenating the RBCs, by incubation at 37°C for 2 hours in the presence of inosine and pyruvate, reduced the supernatant potassium concentration by approximately 10 mmol/L, caused by reuptake of a fraction of the leaked potassium. This reduction of approximately 10 mmol/L was consistent irrespective of storage time, and on Day 4 warming the RBCs reduced the supernatant potassium from 43.1 to 33.1 mmol/L (Fig. 2B), a potassium concentration more usually found on Day 21 of storage (Fig. 3A). Potassium levels in standard units of RBCs prepared for transfusion were measured and the increase during storage plotted (Fig. 3A). From the data, the supernatant of Day 5 stored RBCs in SAG-M would be estimated to contain a mean of approximately 11 to 12 mmol/L Volume **, ** **
TRANSFUSION
3
BAWAZIR ET AL.
A
B
22
65
20
60
K in supernatant (mmol/L)
18
K in plasma (mmol/L)
Population frequency of the Arg723Gln SNP
70
16 14 12 10 8
55 50 45 40 35 30 25 20 15
6
10
4
5 0
10 20 30 40 50 60 70 80
Time (hr)
0
2
4
6
8
10
12
14
Time (days)
Fig. 2. Cation permeability of RBCs from the London blood donor. (A) A heparinanticoagulated blood sample from the London donor was incubated over 3 days at
The Arg723Gln substitution in ABCB6 is listed on dbSNP with identifier rs148211042 but is not linked to a disease or phenotype. The 1000 Genomes, UK10K, NHLBI-ESP, and the ClinSeq projects all provide frequency data for rs14811042. Overall the data show that this SNP is present at an allele frequency of approximately 1:1000 in European populations (22 in 17,875 alleles observed [Table 1]) such that the heterozygous state will be present in approximately 1 in 500 European blood donors (95% CI, 1 in 270 to 1 in 625).
DISCUSSION
RBC membrane cation leaks that are temperature dependent have been RT (open symbols) or 4°C (closed symbols) and the plasma potassium concentration described in a variety of disorders was measured by flame photometry. (▽) Control; (▼) control 4°C; (○) London donor; ranging from moderately severe hemo(●) London donor 4°C. (B) A unit of leukoreduced RBCs in SAG-M from the London lytic states to asymptomatic FP.7 donor was stored for 2 weeks at 4°C. Samples were taken on Days 1, 4, 7, and 14 and The disorders are dominantly inherited either left at 4°C (untreated) or incubated at 37°C for 2 hours in the presence of with complete penetrance. Clinically inosine and pyruvate (rejuvenated). Extracellular potassium concentration was measignificant forms include Southeast sured by flame photometry. (○) London donor untreated; (●) London donor rejuveAsian ovalocytosis,12 dehydrated and nated. Error bars show the SD from the mean (n = 3). overhydrated stomatocytosis, and cryohydrocytosis.7,13 Cases with a clinical phenotype are generally very rare in European popupotassium (Fig. 3A). This is consistent with that measured lations and will often render an affected individual in units of RBCs stored in CPD plasma on Day 5 where unsuitable for blood donation due to anemia or other median supernatant potassium levels were 11.94 mmol/L problems. FP, which has been described in a number of (range, 6.00-21.60 mmol/L; n = 61). The data from Day 7 kindreds, is clinically benign and probably underdiagon 177 units shows a skewed distribution with a median nosed. Asymptomatic individuals with FP may therefore value of 14.0 mmol/L and 2.5th and 97.5th percentiles of become blood donors. Several different forms of FP have 6.1 and 31.8 mmol/L, respectively (Fig. 3B). The units of been described with cation leaks that vary in their temRBCs in the two case reports described here were on Day 5 perature dependence (illustrated in Fig. 4).7 For instance or Day 6 of storage and the supernatant potassium in both cases (33.9 and 41.4 mmol/L, respectively) was well above FP-Edinburgh is a dehydrated stomatocytosis or the expected level for RBCs at that time in storage and xerocytosis type, caused by a mutation in PIEZO114 that above the 97.5th percentile even for the data set on Day 7 has a minor cation leak at RT but does not leak at 4°C (Fig. 3B). (Fig. 4A).7 The FP-Chiswick type, which includes the famiGenomic DNA was analyzed from the FP-Cardiff lies FP-Falkirk, FP-East London (Bow), and FP-Lille, family and from the two blood donors. The NM_005689.2: caused by the substitutions Arg375Trp and Arg375Gln in c.2168G>A mutation leading to the substitution ABCB6 reported by Andolfo and colleagues,6 leaks cations p.Arg723Gln was found in the proband of the FP-Cardiff both at 20°C and at 4°C (Fig. 4B).7,15 However, neither family (I.1; Fig. 1). DNA analysis of an additional seven Arg375 mutation is listed in the dbSNP, and these mutamembers of the FP-Cardiff family showed that the tions are probably very rare. The FP-Cardiff type described NM_005689.2:c.2168G>A mutation segregates with FP in here has RBC cation permeability that is normal at 37°C, the Cardiff family. The three affected FP-Cardiff family only mildly abnormal at 20°C, but at 4°C it is six times that members were heterozygous for the substitution and the of control RBCs (Fig. 4C).7,10 five unaffected members expressed only Arg723 (Fig. 1). In this study, we describe a novel FP mutation in the The same mutation and p.Arg723Gln substitution was also ABCB6 gene that cosegregates with FP in a family from found in both blood donors. Cardiff (FP-Cardiff) and results in the heterozygous 4
TRANSFUSION Volume **, ** **
FAMILIAL PSEUDOHYPERKALEMIA IN BLOOD DONORS
70 60 50 40 30 20 10
40 30 20
Mean (15.5)
42 ay
D
ay
35
28 D
ay
21
2.5% Median (6.1) (14.0)
D
ay
14 D
ay D
ay
1
50
D
D
Number of un its
B
7
0
ay
Supern natant Potassiu um (mmol/L)
A
97.5% (31.8)
Case 1 Day 6 33mmol/L Case 2 Day 5 41mmol/L
10
4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42
0
Supernatant potassium (mmol/L) Fig. 3. Supernatant potassium concentration of stored donor RBC units. (A) Supernatant potassium concentration of RBC units over 42 days. The supernatant potassium of RBCs is routinely monitored for quality by the Components Development Laboratory (CDL) NHSBT. Data presented were measured over 42 days of refrigerated storage in randomly selected units between 2003 and 2006. The data for each time point are represented by the median (solid bar), the 25th and 75th percentiles of the median (box), and the 2.5th and 97.5th percentiles of the median (error bars). Day 1, n = 538; Day 7, n = 177; Day 14, n = 137; Day 21, n = 69; Day 28, n = 29; Day 35, n = 173; Day 42, n = 119. (B) Supernatant potassium concentration of RBC units on Day 7 of storage. The graph shows the distribution of the potassium concentration in RBC units on Day 7 of refrigerated storage as collected by CDL, NHSBT. Data represent randomly selected units collected between 2003 and 2006 (n = 177). The distribution of the data is not normal but is right-skewed. Both the median and median percentile data and the mean data are shown above the bar chart. The potassium concentration of each of the FP-Cardiff type donor units reported in this article was measured on Day 5 or Day 6 so are not directly comparable to the Day 7 CDL data; however, both these units would fall within the right-hand tail of the distribution at positions indicated by arrows.
substitution (Arg723Gln) in ABCB6 protein. The same SNP was found in the two blood donors with unusually high potassium concentrations (33.9 and 41.4 mmol/L) in recently stored donations. Pseudohyperkalemia was confirmed in the donors; their blood plasma potassium after storage was only slightly raised at 20°C but increased much more rapidly than the control when stored at 4°C identical to the changes previously recorded for FP-Cardiff.10 Unlike other described mutations in ABCB66 and PIEZO114 causing FP, this newly described mutation is relatively common with a frequency of this dominant trait of approximately 1 in 500 in persons of European origin. Potassium in the supernatant of stored cells increases during storage (Fig. 3A). For the majority of transfusions this is probably of limited significance because the total amount of potassium transfused is relatively small compared to the total blood volume of the recipient. However, there are certain clinical scenarios where transfusion of high-potassium donor blood may be more of a concern. In large-volume neonatal and pediatric transfusion, high levels of supernatant potassium can have serious or fatal consequences.16-20 The level of transfused supernatant potassium that could be potentially dangerous to a recipient would depend on the rate of transfusion, the relative volume of transfusion compared to the neonatal or infant blood volume, and other contributory factors such as the clinical status of the recipient (recently reviewed by Lee et al.20). For example, a study by Brown and colleagues21 suggested in a simulated model that low cardiac output states can be associated with hyperkalemia if the potassium concentration in the supernatant of the rapidly transfused blood exceeds a level as low as 10 mmol/L. It is common for blood services to restrict the shelf life of RBCs for large-volume neonatal transfusion to reduce the risk of hyperkalemia, and some hospitals may wash RBCs to reduce potassium levels before transfusion. Most hospital departments in the United Kingdom using blood for neonatal bypass prime measure the potassium of the prime fluid before use, although perfusion practice is variable.22 However, blood for large-volume transfusion in neonates in other circumstances, although stored for a short period only, is not routinely tested for supernatant potassium. The RBC unit from the London donor described here contained 17.5 mmol/L potassium on Day 1 of storage, and the potassium level rapidly increased to 43.1 mmol/L by Day 4 of storage, a level commonly found on Day 35 of storage in standard RBC units. After rejuvenation of the RBCs we found some reduction in supernatant potassium in the FP donor unit. For example, after 4 days of storage at 4°C rejuvenation reduced the supernatant potassium from 43.1 to 33.1 mmol/L, but this is a level that would still be considered potentially dangerous for a large-volume transfusion to a child. RBC units with unexpectedly high potassium Volume **, ** **
TRANSFUSION
5
BAWAZIR ET AL.
TABLE 1. Database analysis of allele frequency* European (allele frequency) 9:8,591 1:757 10:7,500 2:1,027 22:17,875 (0.1%)
Database NHLBI Exome Sequencing Project 1000 Genomes project UK10K project CSAgilent ClinSeq project Total
African American (allele frequency) 2:4404
0.14
C
0.12 0.10 0.08 0.06 0.04 0.02 0.00 40
30
20
10
Temperature (°C)
0
Ouabain+bumetanide K influx (mmol/L cells/h)
B Ouabain+bumetanide K influx (mmol/L cells/h)
A
Ouabain+bumetanide K influx (mmmol/L cells/h)
* Frequency data were available from four databases, 1000 Genomes (1092 individuals), the NHLBI Exome Sequencing Project (2203 African American and 4300 European Americans), the ClinSeq project (514 individuals of European descent), and the UK10K wholegenome sequencing project (3621 individuals from the UK). Overall the data suggest an allele frequency of 1:1000. As the heterozygous p.R723Q (c.g2168a) mutation is dominantly inherited with complete penetrance this suggests that this mutation may be present in 1:500 individuals of European descent.
0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 40
30
20
10
Temperature (°C)
0
0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 40
30
20
10
0
Temperature (°C)
Fig. 4. Illustration showing the temperature dependence of the cation leak in different FP phenotypes. These data are shown only as an illustration and are redrawn from Stewart.7 The method measured potassium influx using 86Rb as a tracer in the presence of 0.1 mmol/L ouabain and bumetanide as described.10 The data illustrate the difference in the temperature dependence of the cation leak between the three FP phenotypes. FP-Cardiff RBCs are unique in that their cation permeability is almost identical to control RBCs at 37°C but is six times normal at 0°C. (A) Temperature dependence of potassium flux in FP-Edinburgh RBCs. (○) Control; (●) FP-Edinburgh. (B) Temperature dependence of potassium flux in FP-Chiswick RBCs. (○) Control; (●) FP-Chiswick. (C) Temperature dependence of potassium flux in FP-Cardiff RBCs. (○) Control; (●) FP-Cardiff.
from donors such as we have described could potentially cause significant hyperkalemia in these patients.20,23,24 There is increasing awareness that inherited changes in RBCs or platelets (PLTs) may affect survival and function of transfused cells.25 Here, we have described a significant RBC abnormality due to a relatively common polymorphism that causes no known clinical problems in the donor. Potassium leaks may be detected in affected individuals when unexpected hyperkalemia occurs in blood samples kept overnight at RT before testing. However, when hyperkalemia is detected in such patients, the plasma potassium is typically repeated urgently on a fresh sample and found to be normal. The original abnormal potassium is then dismissed as a laboratory error or 6
TRANSFUSION Volume **, ** **
attributed to lysis. Therefore, these donors can go unrecognized both clinically as patients and as blood donors. Standard RBC units stored for up to 35 days in the United Kingdom, or 42 days elsewhere, may have supernatant potassium levels of 50 mmol/L but total potassium loads of only approximately 5 mmol/unit due to the relatively small volume of the supernatant. It is to be expected that there will be variation in the level of potassium in the supernatant of RBC units, determined by factors related both to the donor and to changes in the RBCs due to processing and storage; however, published data on this distribution in large data sets are lacking. We have shown that in standard RBC units during storage there is a wide range of values of supernatant potassium. Supernatant
FAMILIAL PSEUDOHYPERKALEMIA IN BLOOD DONORS
potassium concentration in units stored for 7 days mainly fall within the 2.5th to 97.5th percentile range (6.131.8 mmol/L) but there is a tail end of units that have levels above the 97.5 percentile from the median (Fig. 3B). Our data demonstrate that donors with the FP mutation may contribute to the latter data on Day 5 or 6 of storage. However, potassium levels in RBC units from these donors will not be easily distinguishable from those of normal donors during the later weeks of normal RBC storage when potassium levels reach a plateau (Fig. 3A; Fig. S1, available as supporting information in the online version of this paper). The clinical significance of the FP mutation for RBC transfusion is likely to be greatest for large-volume transfusions to neonates, where units are deliberately selected to be no more than 5 days postdonation, as units from donors with the mutation are likely to have potassium levels that are well above average for the unit age. However, it is possible that there could be additional effects, such as altered RBC functionality or decreased stability of the RBC membrane causing increased vesiculation, that could occur with increasing lengths of storage in RBCs units from donors with FP-Cardiff and this is an area of ongoing investigation. Having identified the FP-Cardiff mutation and highlighted the potential clinical significance, it will be important to determine the genotype and frequency of other FP causing mutations. The development of a method that could reliably identify those individuals bearing the common FP mutations in the donor population could be considered alongside other genomic donor screening tests for blood groups, HLA types, and other genetic variations as an effective way of improving the quality of the RBC supply and the provision of highly selected units for specific patient groups in the future. In conclusion, we report here a novel mutation that may affect 1:500 clinically asymptomatic European blood donors and that leads to an increased rate of potassium leakage from RBCs when stored at 4°C. Recognition and testing for genetic variants affecting RBC or PLT transfusion physiology such as described here could help improve the quality of the blood supply. We are currently undertaking a series of further studies to fully elucidate the implications of this finding for RBC transfusion and to inform policy on how to manage donors who may be identified with the FP mutation.
research design; and all authors contributed to the manuscript preparation.
WEB-BASED RESOURCES (ACCESSED IN AUGUST 2013) dbSNP rel 137 (http://bioq.saclab.net/query/submit.php ?db=bioq_dbsnp_human_137) dbSNP (http://www.ncbi.nlm.nih.gov/snp). 1000 Genomes (http://browser.1000genomes.org). NHLBI Exome Sequencing Project (http://evs.gs .washington.edu/). ClinSeq project (http://www.genome.gov/20519355). UK10K whole-genome sequencing project (http:// www.uk10k.org/). CONFLICT OF INTEREST NHSBT has filed a patent “Method of detecting cation leaks” Patent No GB1219226.6. Apart from the patent, the authors have disclosed no conflicts of interest.
REFERENCES 1. Hess JR. Red cell changes during storage. Transfus Apher Sci 2010;43:51-9. 2. Bruce LJ, Robinson HC, Guizouarn H, et al. Monovalent cation leaks in human red cells caused by single aminoacid substitutions in the transport domain of the band 3 chloride-bicarbonate exchanger, AE1. Nat Genet 2005;37: 1258-63. 3. Bruce LJ, Guizouarn H, Burton NM, et al. The monovalent cation leak in overhydrated stomatocytic red blood cells results from amino acid substitutions in the Rh-associated glycoprotein. Blood 2009;113:1350-7. 4. Flatt JF, Guizouarn H, Burton NM, et al. Stomatin-deficient cryohydrocytosis results from mutations in SLC2A1: a
5.
6.
7.
8. ACKNOWLEDGMENTS We thank J. Hosken for DNA sequencing. We thank the donors, patients, and their families for their cooperation. WMB performed DNA and cation analysis; JFF, MW, and RAC performed cation analysis; AR performed population genetics analyses; CEC, LP, and HVN carried out provision of donor samples; GWS carried out provision of patient samples; LJB and JPW performed
9.
10.
novel form of GLUT1 deficiency syndrome. Blood 2011; 118:5267-77. Zarychanski R, Schulz VP, Houston BL, et al. Mutations in the mechanotransduction protein PIEZO1 are associated with hereditary xerocytosis. Blood 2012;120:1908-15. Andolfo I, Alper SL, Delaunay J, et al. Missense mutations in the ABCB6 transporter cause dominant familial pseudohyperkalemia. Am J Hematol 2013;88:66-72. Stewart GW. Hemolytic disease due to membrane ion channel disorders. Curr Opin Hematol 2004;11: 244-50. Stewart GW, Corrall RJ, Fyffe JA, et al. Familial pseudohyperkalaemia. A new syndrome. Lancet 1979;2: 175-7. Helias V, Saison C, Ballif BA, et al. ABCB6 is dispensable for erythropoiesis and specifies the new blood group system Langereis. Nat Genet 2012;44:170-3. Gore DM, Chetty MC, Fisher J, et al. Familial pseudohyperkalaemia Cardiff: a mild version of cryohydrocytosis. Br J Haematol 2002;117:212-4. Volume **, ** **
TRANSFUSION
7
BAWAZIR ET AL.
11. National Blood Service. Guidelines for the blood transfusion services in the UK. 8th ed. London, UK: The Stationery Office; 2013. [cited 2013 Aug]. Available from: http:// www.transfusionguidelines.org.uk/red-book 12. Guizouarn H, Borgese F, Gabillat N, et al. South-east Asian ovalocytosis and the cryohydrocytosis form of hereditary stomatocytosis show virtually indistinguishable cation permeability defects. Br J Haematol 2011;152:655-64. 13. Delaunay J. The hereditary stomatocytoses: genetic disorders of the red cell membrane permeability to monovalent cations. Semin Hematol 2004;41:165-72. 14. Andolfo I, Alper SL, De Franceschi L, et al. Multiple clinical forms of dehydrated hereditary stomatocytosis arise from
23. Vraets A, Lin Y, Callum JL. Transfusion-associated hyperkalemia. Transfus Med Rev 2011;25:184-96. 24. Strauss RG. RBC storage and avoiding hyperkalemia from transfusions to neonates & infants. Transfusion 2010;50: 1862-5. 25. Kleinman S, Busch MP, Murphy EL, et al. The National Heart, Lung, and Blood Institute Recipient Epidemiology and Donor Evaluation Study (REDS-III). The National Heart, Lung, and Blood Institute Recipient Epidemiology and Donor Evaluation Study (REDS-III): a research program striving to improve blood donor and transfusion recipient outcomes. Transfusion 2013;54(3 Pt 2):942-55.
mutations in PIEZO1. Blood 2013;121:3925-35. 15. Carella M, d’Adamo AP, Grootenboer-Mignot S, et al. A second locus mapping to 2q35-36 for familial pseudohyperkalaemia. Eur J Hum Genet 2004;12:1073-6. 16. Hall TL, Barnes A, Miller JR, et al. Neonatal mortality following transfusion of red cells with high plasma potassium levels. Transfusion 1993;33:606-9. 17. Chen CH, Hong CL, Kau YC, et al. Fatal hyperkalemia during rapid and massive blood transfusion in a child undergoing hip surgery—a case report. Acta Anaesthesiol Sin 1999;37:163-6. 18. Baz EM, Kanazi GE, Mahfouz RA, et al. An unusual case of hyperkalaemia-induced cardiac arrest in a paediatric patient during transfusion of a “fresh” 6-day-old blood unit. Transfus Med 2002;12:383-6. 19. Smith HM, Farrow SJ, Ackerman JD, et al. Cardiac arrests associated with hyperkalemia during red blood cell transfusion: a case series. Anesth Analg 2008;106:1062-9. 20. Lee AC, Reduque LL, Luban NL, et al. Transfusionassociated hyperkalemic cardiac arrest in pediatric patients receiving massive transfusion. Transfusion 2014; 54:244-54. 21. Brown KA, Bissonnette B, McIntyre B. Hyperkalaemia during rapid blood transfusion and hypovolaemic cardiac arrest in children. Can J Anaesth 1990;37:747-54. 22. Harvey B, Shann KG, Fitzgerald D, et al.; American Society of ExtraCorporeal Technology’s International Consortium for Evidence-Based Perfusion and Pediatric Perfusion Committee. International pediatric perfusion practice: 2011 survey results. J Extra Corpor Technol 2012;44: 186-93.
8
TRANSFUSION Volume **, ** **
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article at the publisher’s Web site: Fig. S1. Distribution of supernatant potassium concentration of stored donor red cell units. The graphs show the distribution of the potassium concentration in red cell units at weekly intervals of refrigerated storage as collected by CDL, NHSBT. Data represents randomly selected units collected between 2003-2006. The distribution of the data varies with the week of storage. The median and median percentile data are shown above the bar chart. The potassium concentration of each of the FP-Cardiff type donor units reported in this paper was measured on Day 5 or Day 6 and are shown on the Day 7 graphs; both these units fall within the righthand tail of the distribution at positions indicated by arrows. A) Distribution of the supernatant potassium concentration of red cell units on Day 0/Day 1 of storage. B) Distribution of the supernatant potassium concentration of red cell units on Day 7 of storage. C) Distribution of the supernatant potassium concentration of red cell units on Day 14 of storage. D) Distribution of the supernatant potassium concentration of red cell units on Day 21 of storage. E) Distribution of the supernatant potassium concentration of red cell units on Day 35 of storage. F) Distribution of the supernatant potassium concentration of red cell units on Day 42 of storage.
