Vox Sanguinis (2005) 88, 207–221

© 2005 Blackwell Publishing

INTERNATIONAL FORUM

Transfusion of apheresis platelets

Blackwell Publishing, Ltd.

and ABO groups

Several incidents of intravascular haemolysis caused by ABO In several countries/centres, haemolysis has been reported
antibodies have been reported after the transfusion of apheresis after the transfusion of apheresis platelets across a minor ABO
platelets across a minor ABO incompatibility. The relatively incompatibility. It should be appreciated that such incidents
large volume of plasma in concentrates of apheresis platelets have become rare because measures are now generally taken
increases the risk of this complication. It therefore seemed of to prevent this complication (see below). To illustrate this
interest to acquire information on the occurrence of this com- point, in the USA prior to 1990, when platelets from ABO-
plication as well as on the measures that are taken to prevent it. mismatched donors were routinely administered, patients
To obtain this information, the following questions were with a positive direct antiglobulin test on the red cells, an
sent to experts in the field. We obtained 16 contributions to increased need of red cell transfusions and other evidence of
this Forum: haemolysis, were frequently seen. Since then, platelets from
Question 1. Have you seen cases of haemolysis after trans- group O donors are rarely given to non-O recipients and inci-
fusion of apheresis platelets or, perhaps, after transfusion dents of posttransfusion haemolysis no longer occur. It is also
of pooled concentrates? suggested that there is insufficient awareness of this problem
Question 2. Do you, in your country/centre, take measures to and that such incidents are under-reported (Kretschmer).
prevent haemolysis due to anti-A/B in platelet concentrates No incidents of intravascular haemolysis have been
in recipients of apheresis platelets, e.g. in case HLA-, or reported to occur after the transfusion of pooled platelet
HPA-matched platelets are required? concentrates.
Question 3. If you take measures to prevent haemolysis in In all countries/centres, in principle, platelets from ABO-
recipients due to anti-A/anti-B in platelet concentrates, identical donors are used to prevent the accelerated destruc-
which of the following have you adopted: tion of ABO-incompatible platelets in the recipient as well as
a) Only platelets from ABO-identical donors are transfused. the destruction of red cells owing to the transfusion of ABO-
b) In case platelet concentrates are not ABO identical, do incompatible plasma. If, because human leucocyte antigen
you determine the titre of IgM and IgG anti-A/B and do (HLA)-, or human platelet antigen (HPA)-compatible platelets
you exclude donors with titres above a critical level? are required and ABO-identical donors are not available, the
If you follow this policy: transfusion of apheresis platelets across a minor ABO incom-
• which technique(s) do you use to determine the titres; patibility cannot be avoided, measures are taken in all countries/
• which titres do you consider to be critical; centres to prevent haemolysis in the recipients (see below). In
• do you permanently exclude donors with titres above the such cases, the use of platelets from A donors for B recipients,
critical level from the donor panel for transfusions to and vice versa, is preferred over the use of O donors for
recipients whose red cells are incompatible with anti-A/B recipients of another ABO blood group. Measures taken are
in the concentrate; and/or as follows:
• do you prefer donors with blood group A or B over • the titre of anti-A/B in the donor is determined and only
donors with blood group O? platelets from donors with low titres are used (see below);
c) Do you resolve the problem by reducing the volume of • the amount of plasma is reduced to ≈ 90 ml;
plasma? • the plasma is replaced by platelet additive solution;
d) Do you replace the plasma by platelet additive solution? • the incompatible plasma is replaced by AB plasma; and/or
e) Other measures? • the platelets are washed and resuspended in saline.
Question 4. Since it has been clearly shown that anti-A/B in Techniques used for determining the titre of anti-A/B are:
the recipient can considerably shorten the lifespan of incom- tube saline agglutination; tube saline agglutination followed
patible platelets, do you use ABO-compatible platelets: by an indirect antiglobulin test; the micro column agglutina-
• for all patients; tion technique, an Olympus PK7200; and – in two centres –
• for selected groups, e.g. those with strong anti-A/B; determining whether the alloantibodies are haemolytic in
• or do you not take ABO groups into account? vitro or by actually determining the haemolysin titre.

207
208 International Forum

Values considered to be critical are titres of > 1 : 64 to H. W. Reesink

1 : 100 for immunoglobulin M (IgM) and > 1 : 256 to 1 : 400 Sanquin Blood Bank North-West Region and Sanquin
for immnoglobulin G (IgG) and of > 1 : 16 for haemolysins Diagnostic Services
or just when haemolysis occurs in vitro. Evidence that the PO Box 9137
NL-1006 AC Amsterdam
titre of the alloantibodies in the donor is indeed critical with
the Netherlands
regard to the occurrence of haemolysis in the recipient, is
E-mail: [email protected]
presented by Reinhardt: no haemolysis was found to occur I88
International
nternational
3April 2005 Forum

with concentrates from donors with a titre of ≤ 1 : 64,

whereas severe haemolysis occurred in a group A patient
K. G. Davis, J. Lown & A. Thomson
transfused with an apheresis concentrate from a group O
donor with an anti-A titre of > 1 : 512. Question 1
Donors with high titres of anti-A/B are not excluded from We know of at least two documented incidents of intravas-
the donor panel, but their platelets are not used across a minor cular haemolysis (one involving an infant and the other an
ABO incompatibility. Heal and Blumberg draw attention to adult) after transfusion of an apheresis product. We have
the fact that, in addition to accelerated red cell destruction, heard of other anecdotal cases, but they are uncommon. Most
there are other disadvantages of transfusing ABO-incompatible concern comes during pediatric use when cross-grouping is
plasma, such as the formation of immune complexes which necessary.
bind and activate complement and which then bind to plate-
lets, leading to phagocytosis by monocytes/macrophages and Question 2
carrying unknown potential for morbidity e.g. proinflam- At present there is varying practice in Australia with respect to
matory predisposition for multi-organ failure and death in the measures undertaken to reduce the possibility of haemolysis.
cardiac surgery. A similar remark is made by Kretschmer. In some jurisdictions, buffy coat-pooled platelets are plasma
In all countries/centres, if HLA- or HPA-compatible plate- reduced and resuspended in ‘T-Sol'. In another jurisdiction,
lets are required and no ABO-identical donors are available, 100% of platelets are obtained from apheresis and are screened
ABO incompatibility is preferred over HLA or HPA incompat- for ‘high-titre' donor antibodies.
ibility, one of the above measures being taken to prevent the
problems attached to the transfusion of ABO-incompatible Question 3
plasma in the platelet concentrate. However, in one centre a) See the response to question 4.
(Novotny & Brand), when the post-transfusion platelet incre- b) Yes in some jurisdictions. The following is an example
ment is insufficient, the titre of anti-A/B in the recipient is of a method being used – sal RT progressing to the indirect
determined, and if found to be 1 : 128 to 1 : 256, only ABO- antiglobulin test (IAT) for all apheresis units; a one-off
compatible platelets are used. dilution of 1 : 200 used, if negative = low titre; high-titre
In conclusion, there is general awareness of the danger donors are not permanently excluded. Group A donors have
of haemolysis (and other problems) after the transfusion of been used instead of O for group B patients.
apheresis platelets containing ABO-incompatible plasma. c) As stated above in the response to question 2, some
Measures to prevent this problem are taken universally, jurisdictions produce plasma-reduced buffy coat platelet
which explains why this complication now rarely occurs. pools to avoid the haemolysis issue.
It goes without saying that the reader is advised to read the d) Yes, T-Sol is used as a plasma replacement in buffy coat
individual answers, which contain much further interesting pools.
information. e) Currently a process is underway to validate T-Sol-
suspended apheresis platelets. The long-term desire is to obtain
R. N. I. Pietersz 100% plasma-depleted apheresis platelets in additive solution.
Sanquin Blood Bank North-West Region
PO Box 9137 Question 4
NL-1006 AC Amsterdam Consideration is given to ABO (and in some instances RhD)
the Netherlands
groups. The following is an extract from national guidelines [1]:
E-mail: [email protected]
Platelet concentrates, in order of preference, should be:
(i) Patient's own ABO, Rh(D) group.
C. P. Engelfriet
Sanquin Diagnostic Services and Sanquin Research If this is not possible, a decision on whether to give antigen
PO Box 9190 or plasma-incompatible platelets may be of importance,
NL-1006 AD Amsterdam depending on the patient diagnosis/therapy:
the Netherlands (ii) ABO, Rh(D) antigen compatible (but plasma incom-
E-mail: [email protected] patible), or

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

 International Forum 209

(iii) ABO, Rh(D) antigen incompatible. If these are not available, we take measures to reduce
Requirements for HLA compatibility may take precedence exposure to incompatible plasma, especially in children.
over ABO typing.
Individual units of different ABO blood groups shall not Question 3
be pooled. Matching for Rh(D) type is desirable (as platelet a) The majority (≈ 85 %), but not all, patients are transfused
products may contain small or minimal numbers of red cells), with ABO-identical platelets, especially when apheresis plate-
but may be less important than ABO matching. Platelets do lets are used. In a group of bone marrow transplant patients
not carry Rh antigens. with graft/recipient ABO mismatch, we respect the ABO group
The administration of Rh(D) immunoglobulin should be of the donor/graft.
considered for Rh(D)-negative patients, especially premeno- b) We test all (A, B and O) apheresis donors for anti-A/B by
pausal females, when platelet concentrates from Rh(D)- using the agglutination test in saline (e.g. for immunoglobulin
positive donors are transfused. M); titres are written on the label of the product. Platelets
from donors with a titre of > 1 : 64 are used exclusively for
Reference ABO-identical recipients.
1 Scientific Subcommittee: Guidelines for Pretransfusion Test- In theory, when the donor’s anti-A/B titre is not known,
ing, edn 4. 2002. Sydney, Australian & New Zealand Society we prefer to use donors with blood group A or B rather than
of Blood Transfusion (www.anzsbt.org.au) donors with blood group O. In practice we do not usually
have to make this choice owing to the above-mentioned
K. G. Davis
procedures.
Chief Medical Scientist
About 25% of apheresis platelet concentrates in our centre
Transfusion Medicine Unit
Institute of Medical & Veterinary Science
are produced/delivered as ‘platelets in additive solution’. We
Royal Adelaide Hospital reduce plasma volume only in special situations (for small
Adelaide babies, etc.).
South Australia
E-mail: [email protected] Question 4
We prefer ABO-identical platelets, except when required for
J. Lown refractory or immunized patients where human leucocyte
Principal Scientist antigen (HLA) or human platelet antigen (HPA) platelets
Transfusion Medicine Unit are required (5). If this is not possible, we respect ‘plasma
Royal Perth Hospital
compatibility’, especially in children, or we reduce the
Perth
plasma content, replacing plasma by additive solution. The
Western Australia
anti-A/B titre in the recipient is taken into account only in
E-mail: [email protected]
patients after mismatch bone marrow transplantation.
A Thomson In the worst-case scenario (vital indications) we use buffy
Consultant Hematologist coat-derived platelets of different blood groups (e.g. a set of
Royal North Shore Hospital individual bags of different ABO groups, some of which could
Sydney be ABO incompatible).
NSW
Australia
E-mail: [email protected]
References
3IApril
88
International
nternational
2005 Forum 1 Murphy MF, Hook S, Waters AH, Sterlini J, Whelan J, Davis C,
Lister TA: Acute haemolysis after ABO-incompatible platelet
transfusions. Lancet 1990; 21;335: 974–795
P. Turek
2 Shanwell A, Ringden O, Wiechel B, Rumin S, Akerblom O: A
Question 1 study of the effect of ABO incompatible plasma in platelet
No severe haemolysis caused by ABO incompatibility after concentrates transfused to bone marrow transplant recipi-
ents. Vox Sang 1991; 60:23–27
administration of platelet concentrate has been observed
3 Mair B, Benson K: Evaluation of changes in hemoglobin
during (at least) the last 10 years in our center, but a risk of
levels associated with ABO-incompatible plasma in apheresis
haemolysis is well known from the literature (1,2,3,4).
platelets. Transfusion 1998; 38:51–55
4 Duguid JK, Minards J, Bolton-Maggs PH: Lesson of the
Question 2 week: incompatible plasma transfusions and haemolysis in
To prevent haemolysis we prefer to use ABO-identical or children. BMJ 1999; 16;318:176–177
at least ABO plasma-compatible apheresis platelets in our 5 Petz LD, Garratty G, Calhoun L, Clark BD, Terasaki PI, Gresens C,
centre (a similar policy is adhered to throughout the country). Gornbein JA, Landaw EM, Smith R, Cecka JM: Selecting

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

210 International Forum

donors of platelets for refractory patients on the basis of All platelet products that are transfused in Finland are
HLA antibody specificity. Transfusion 2000;40:1446–1456 manufactured and supplied to the hospitals by The Finnish
Red Cross Blood Service. Ninety eight per cent of platelet
Petr Turek components distributed to hospitals are pooled products,
Institute of Haematology and Blood Transfusion
which are prepared by the buffy coat method from four
U nemocnice 1
whole-blood units and suspended in platelet additive
128 20 Prague
Czech Republic
solution. Plasma accounts for less than 20% of the volume
E-mail: [email protected] of the pooled platelet products, which means that ≈ 50 ml
I88
International
nternational
3April 2005 Forum
of plasma is present in a four-donor platelet component.
Two per cent of platelet components are single-donor
E. Dickmeiss products, which are prepared by apheresis and suspended in
donor plasma. Plasma accounts for most of the volume of
Question 1 apheresis platelet products, which means that ≈ 190 ml
Minor ABO-incompatible apheresis platelet transfusions are of plasma is present in an apheresis platelet component
avoided, and we have not seen haemolysis after transfusion corresponding to four platelet units. Apheresis platelets are
with minor incompatible buffy coat-derived platelets pooled transfused almost exclusively to human leucocyte antigen
with additive solution. (HLA)- and/or human platelet antigen (HPA)-alloimmunized
patients.
Question 2
Yes. In recipients of apheresis platelets, we take measures to pre-
Question 1
vent haemolysis caused by anti-A/B in platelet concentrates. The Finnish Red Cross Blood Service collects data on the
adverse effects of the transfusion of blood components in
Question 3 Finland. The collection system is based on voluntary
We replace the plasma with AB plasma if minor ABO incom- reporting from hospitals and there have been no reports of
patibility is unavoidable in human leucocyte antigen (HLA)- haemolytic reactions caused by the transfusion of pooled or
or human platelet antigen (HPA)-matched transfusions. single-donor apheresis platelet products.

Question 4
Question 2
Major ABO-incompatible platelet transfusions are avoided
We are aware of the possibility of haemolysis caused by
for all patients. However, if the only available HLA-matched
donor anti-A/B, especially in recipients of apheresis platelets.
platelets to a given HLA-ABC immunized patient are major
At present we try to prevent haemolysis caused by anti-A/B
ABO incompatible, we prefer to use these in lieu of ABO-
in platelet components primarily by supplying ABO-identical
compatible, but HLA-mismatched, platelets.
apheresis platelet products to hospitals.
We have a donor registry of 10 100 active HLA-typed
Ebbe Dickmeiss
Copenhagen City Blood Transfusion Service
platelet donors, and for most HLA-immunized patients it is
Section 2034, Rigshospitalet possible to find 3/4 or 4/4 HLA AB-matched platelets, which
Blegdamsvej 9 are either ABO identical or ABO compatible. Transfusion of
DK-2100 Copenhagen ABO-compatible apheresis platelets is considered relatively
Denmark safe because it seems that the Finnish donor population has
E-mail: [email protected] quite low titres of anti-A and anti-B. For example, we
3IApril
88
International
nternational
2005 Forum

recently determined anti-A and anti-B titres from 139 blood

group O platelet apheresis donors, in tubes, by using the
direct agglutination method. The median anti-A or anti-B
T. Krusius & J. Matilainen
titre ranged from 1 : 8 to 1 : 16. Only 5·7% of the donors had
The Finnish Red Cross Blood Service is an economically and a titre of > 1 : 32, and none had a titre of > 1 : 128. In gen-
operatively independent section of the Finnish Red Cross. It eral, the titres were clearly lower than seen in haemolytic
is responsible for the national blood programme in Finland. reactions, owing to the presence of anti-A/B in platelet con-
It collects blood from voluntary donors, manufactures blood centrates published in the literature.
components and plasma products, and offers laboratory HPA-typed platelets are mainly transfused to neonatal
services and expert consultations on transfusion medicine alloimmune thrombocytopenia (NAIT) patients. HPA-typed
and organ and stem cell transplantation to hospitals. Its areas platelet products are washed with AB RhD-negative plasma
of expertise include blood group serology, platelet serology, to prevent haemolysis caused by anti-A/B if the donor plasma
haemostasis and tissue typing. is not compatible with the ABO antigens of the NAIT patient.

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

 International Forum 211

Question 3 platelets (donor blood group O, recipient A). In addition,

The Finnish Red Cross Blood Service favours the production several transfusion reactions without recognizable haemolysis
of group O and group A pooled platelets, and patients are have been reported, especially in children, which could only
almost always transfused either with ABO RhD-identical be explained by ABO-minor incompatibility. Following these
platelets or with ABO RhD-compatible platelets.Transfusions experiences, we always administer ABO-identical or only
of pooled platelet components against the donor anti-A/B major-incompatible apheresis platelets. Minor-incompatible
are also considered to be safe owing to the small amount of platelets are only used in the event that antibodies against
plasma in pooled products and also owing to the low titre of human leucocyte antigens (HLA) have to be taken into account
anti-A/B in the Finnish donor population. if the anti-AB titres of the donor are low (see below).
Presently we are discussing how to reduce the potential Pooled platelet concentrates (from four donors) are more
risk of haemolysis in non-ABO identical apheresis platelet liberally chosen by us with regard to ABO-compatibility, as
transfusions. The alternatives are to replace plasma by platelet the plasma volume of every one donor is only ≈ 50 ml and
additive solution or to screen donors for anti-A/B. So far no the risk of including donors with high haemolysin levels is
decision has been made regarding which alternative will be only ≈ 10%. However, we normally restrict the number of
implemented. incompatible platelets to one ABO-minor incompatible pooled
platelet concentrate, except in massive transfusion. Recently,
Question 4 a group A patient developed positive cross-matches with group
In Finland, all patients are transfused with ABO RhD-identical A red cells and a positive direct antiglobulin text (DAT) after
platelets whenever possible. ABO-compatible platelets are receiving two pooled group O platelet concentrates on one
used if identical platelets are not available, and only under day. No haemolysis was reported to our laboratory, probably
exceptional circumstances are ABO-incompatible platelets because massive transfusion followed owing to an emergency
transfused. The anti-A/B status of the patient is not usually revision after surgery to replace a heart valve.
determined prior to platelet transfusions. Patients undergoing
allogeneic stem cell transplantation are transfused only with Question 2
ABO RhD-identical platelets. Exceptionally, ABO-incompatible At our institution, a clear regimen is followed in order to
platelets are transfused to alloimmunized patients, for prevent haemolysis caused by minor-incompatible platelet
example owing to a lack of ABO-compatible 3/4 or 4/4 HLA transfusion, which is explained below (cf. our answer to Ques-
AB-matched platelets. tion 3). In addition, when transfusion reactions in patients
transfused with ABO-minor incompatible platelet concen-
Tom Krusius trates are reported to our laboratory, we always ask the
Medical Director clinicians to determine haemolysis parameters. We believe
E-mail: [email protected]
that many clinical colleagues in Germany, and possibly also
in other countries, are unfortunately insufficiently aware of
Jaakko Matilainen
this problem, a matter which could explain the few reports
Medical Officer
Finnish Red Cross Blood Service on haemolysis caused by the transfusion of ABO-minor
Kivihaantie 7 incompatible platelet concentrates.
00310 Helsinki
Finland Question 3
E-mail: [email protected] In most instances, we transfuse apheresis platelets from ABO-
I88
International
nternational
3April 2005 Forum

identical donors. To determine whether platelet apheresis

concentrates are minor-incompatible, we determine the
titre of immunoglobulin M (IgM) anti-A/B by using a simple
V. Kretschmer & R. Karger
agglutination tube test. Previously, we have already excluded
Question 1 units for minor-incompatible transfusion if titres were deter-
The current German regulations, as a rule, require ABO- mined to be > 1 : 32. As no transfusion reactions have been
compatible platelet transfusions to be performed. The passive reported for several years, we accept titres of < 1 : 100 as a
transmission of alloantibodies with the plasma has to be critical threshold without obtaining any negative feedback.
considered only in special cases. Thus, the transfusion of ABO We do not believe that anti-A/B immunoglobulin G (IgG)
non-identical platelets is permitted and, hence, the transfu- plays a significant role in minor ABO-incompatible platelet
sion of ABO-incompatible platelet concentrates, especially of transfusion.
pooled units, is not unusual in Germany. We do not determine the anti-A/B titres in pooled platelets,
Many years ago, mild haemolysis occurred in a patient but do restrict the number of units (see the response to Ques-
transfused with ≈ 300 ml of minor-incompatible apheresis tion 1). We do not exclude donors with high titres of anti-A/

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

212 International Forum

B from donation either for pooled platelet concentrates or for V. Kretschmer

apheresis platelets. R. Karger
It is certainly preferable to use donors with blood group A Institute for Transfusion Medicine and Haemostaseology
or B instead of O in ABO-minor incompatible platelet trans- University Hospital
Conradistrasse
fusion. The problem of ABO-minor incompatibility has only
Marburg
very rarely been solved by reducing the volume of plasma.
D-35033
Under such circumstances we usually exchange the donor Germany
plasma for group AB plasma. E-mail: [email protected]
I88
International
nternational
3April 2005 Forum

Question 4
P. Reinhardt, M. Wiesneth, H. Schrezenmeier & E. Seifried
Our transfusion regimen is designed to take fresh apheresis
platelet concentrates for haematological/oncological patients Question 1
with bone marrow hypoplasia. Therefore, these patients mainly In our institution, only patients with anti-human leucocyte
receive ABO-identical concentrates. Only if HLA has to be antigen (HLA) and/or human platelet antigen (HPA) antibodies,
taken into account are ABO-major- or ABO-minor incompat- and who are refractory to pooled random donor platelet
ible apheresis concentrates used. In ABO-minor incompati- concentrates, receive HLA/HPA-matched platelet concen-
bility, the critical anti-A or anti-B titre is taken into account, trates collected by apheresis. Therefore, platelet apheresis
as described above. If the platelet increment after ABO- products are prepared on individual demand only and not
major incompatible platelets shows no efficacy, we search for pre-emptively.
another donor. Between January 2004 and October 2004, overall 640 apher-
We administer pooled platelet concentrates in acute esis concentrates (mean plasma volume 279 ±12 ml) and
bleeding complications and in peri-operative and/or post- more than 12 000 pooled platelet concentrates (mean volume
traumatic situations. Here, the choice of ABO-compatible 282 ± 25 ml containing additive solution and less than 40%
or ABO-incompatible platelets is often driven by logistic residual donor plasma) were generated and distributed.
considerations, i.e. by our aim to ensure supply. Less than 1% of the pooled platelet concentrates were given
Nevertheless, we would like to remind of reports of increased across an ABO-minor incompatibility, all others were ABO
morbidity and mortality during induction therapy for acute identical. Of the 640 HLA-matched apheresis concentrates,
leukaemia and allogeneic progenitor cell transplantation when 33% were ABO-minor incompatible, 17% were ABO-major
ABO-mismatched platelet transfusion was performed [1–3], incompatible, and 50% were ABO identical.
which could not be explained only by haemolysis. Also, in this Not a single incident of haemolysis was reported for any
respect ABO non-identical platelet concentrates should be of the pooled or apheresis platelet products distributed.
used for long-term platelet therapy only when supply cannot All blood products were leucocyte depleted and all apheresis
otherwise be ensured, but not in order to use up concentrates donors were screened for allohaemagglutinins/haemolysins.
shortly before they become out of date. An alternative is to
use minor-incompatible concentrates with low or reduced Question 2
anti-A/B levels. However, outcome data are still controversial For the selection of HLA/HPA-matched platelet concentrates,
in patients not requiring long-term platelet therapy [4,5]. ABO-identical donors are favoured; however, an HLA/HPA-
match is given preference. To prevent haemolysis, allo-
References haemagglutinins/haemolysins are currently determined in
1 Heal JM, Kenmotsu N, Rowe JM, Blumberg N: A possible sur- the serum of all apheresis donors whose products are ABO-
vival advantage in adults with acute leukemia receiving ABO- incompatible with the intended recipient [1,2]. The presence
identical platelet transfusions. Am J Hematol 1994;45:189–190 of haemolysins or anti-A/B titres of 1 : 128 determines that
2 Benjamin RJ, Antin JH: ABO-incompatible bone-marrow trans- the platelet concentrate is to be used only for ABO-identical
plantation: the transfusion of incompatible plasma may exacer- transfusion or the volume of the plasma has to be reduced
bate regimen-related toxicity. Transfusion 1999;39:1273–1274 and replaced by additive solution.
3 Heal JM, Blumberg N: The second century of ABO: and now for
ABO-identical platelet concentrates are used whenever
something completely different. Transfusion 1999;39:1155–1159
possible and are a must for the substitution of neonates and
4 Blumberg N, Heal JM, Hicks GL, Risher WH: Association of
infants with a body weight of < 25 kg.
ABO-mismatched platelet transfusions with morbidity and
mortality in cardiac surgery. Transfusion 2001;41:790–793
5 Lin Y, Callum JL, Coovadia AS, Murphy PM: Transfusion of Question 3
ABO-nonidentical platelets is not associated with clinical More than 95% of our donors show allohaemagglutinin titres
outcomes in cardiovascular surgery patients. Transfusion of < 1 : 128. We are not aware of complications associated
2002;42:166–172 with titres of < 1 : 128, justifying our cut-off level of > 1 : 64

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

 International Forum 213

for adults [2]. Titres of > 1 : 64, or positive haemolysins in 3 Gulliksson H, Eriksson L, Hogman CF, Payrat JM: Buffy-
the apheresis products, prompt the removal of donor plasma coat-derived platelet concentrates prepared from half-
from the product and substitution with an additive solution, strenght citrate CPD and CPD whole blood units. Vox Sang
i.e. T-Sol [3]. However, there are no studies that clearly define 1995; 68:152–159
4 Herman JH: Apheresis platelet transfusions:does ABO matter?
a critical isohaemagglutinin titre [4].
Transfusion 2004; 44:802–804
The technique used to determine the anti-ABO titre is as
5 British Committee for Standards in Haematology, Blood
follows. Undiluted donor serum and serial twofold dilutions Transfusion: Guidelines for the use of platelet transfusions.
in 0·9% NaCl are incubated for 5 min with test erythrocytes British Journal of Haematology 2003;122:10–23
of blood groups O, A1, A2 and B, at +20 °C. Erythrocyte
P. Reinhardt
agglutination (allohaemagglutinin) is evaluated and
M. Wiesneth
documented after a brief centrifugation. The tubes are then
H. Schrezenmeier
incubated at +37 °C for 30 min, without a spin, and the super- E. Seifried
natant is tested for haemolysis (isohaemolysin) [2]. Red Cross Blood Services Baden-Württemberg – Hessen
In 2004, six out of 640 platelet apheresis concentrates had Institute Ulm
allohaemagglutinin titres of > 1 : 64 and the plasma of these Helmholtzstrasse 10
products was reduced before transfusion, as described D-89081 Ulm
above. No haemolysis was reported to our institute after Germany
transfusion of platelet apheresis products during a time- E-mail: [email protected]
3April 2005 Forum
88
International

period spanning the last 10 years. However, prior to routine

allohaemagglutinin screening, a patient with blood group A
P. Rebulla, N. Greppi, D. Riccardi & F. Morelati
had received a platelet apheresis concentrate of group O,
resulting in severe haemolysis. Retrospective testing revealed Question 1
an anti-A titre of 512 in the platelet product. The occurrence of haemolysis after transfusion of apheresis
In terms of overall apheresis platelet production, there is platelets or pooled concentrates has not been reported to our
no preference of blood group A over O because the HLA- blood transfusion service, from the 44 clinical wards using
match is given preference. Donors with high anti-A/B titres our platelet products, during the time-period January 2000 to
were not permanently excluded from platelet apheresis, but November 2004, as determined by the review of 232 reports
were tested again. of transfusion-untoward effects to platelet products received
during the same time-period. Our local lack of evidence of
Question 4 haemolysis, despite its intrinsic positive meaning, must be
All platelet substitutions (both pooled and apheresis concen- considered together with additional information on our
trates) are intended to be ABO-identical. Allohaemagglutinin platelet support policies, which are summarized below.
titres of the recipients are not determined. Our standard platelet product is obtained by soft centrifu-
If ABO-identical donors are not available, i.e. owing to gation of a pool of five buffy coats diluted in 300 ml of a
high demand and/or the presence of rare blood group or commercial crystalloid platelet additive solution (T-Sol;
multiple anti-HLA antibodies, determination of allohaemag- Baxter, Maurepas, France). Owing to the availability of pooled
glutinins/haemolysins in platelet apheresis products is concentrates prepared from buffy-coats, we use a very limited
performed to prevent haemolysis; ABO-minor incompatibility number of apheresis platelets, which are required only during
is preferred over ABO-major incompatibility [1,5]. occasional shortages. In 2003 we delivered a total of 4146
The strategy described is successfully used also in our platelet products, comprising 4036 (97·3%) platelet pools and
institutes Baden-Baden, Frankfurt, Mannheim and Kassel, 110 (2·7%) apheresis platelets. Similarly to the buffy-coat
which produce more than 3500 apheresis platelet and more pools, our apheresis platelets are suspended in a medium con-
than 55 000 pooled random donor platelet concentrates per sisting, on average, of 200 ml of plasma and 300 ml of T-Sol.
year without reports of haemolytic transfusion reactions. Approximately 60% of our platelets have been white cell-reduced
at the time of production by filtration and are used for selected
References indications including: immunodepressed patients undergoing
1 Klüter H, Salama A: Thrombozytenkonzentrate; in Vorstand
haemopoietic stem cell transplantation; patients with non-
und wissenschaftlicher Beirat der Bundesäztekammer (eds):
haemolytic febrile transfusion reactions; prevention of cytome-
Leitlinien zur Therapie mit Blutkomponenten und Plasma-
derivaten: Deutscher Ärzte-Verlag Köln, 2003:29–47 galovirus (CMV) transmission or reactivation; and administration
2 Josephson CD, Mullis NC, Van Demank C, Hillyer CD: Signific- of human leucocyte antigen (HLA)-compatible platelets.
ant numbers of apheresis-derived group O platelet units have The main features of our platelet support policy, which are
‘high-titre’ anti-A/A,B: Implications for transfusion policy. relevant to the discussion on the occurrence of haemolysis,
Transfusion 2004; 44:805–808 are as follows:

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

214 International Forum

(1) The prevalent use of pooled concentrates as compared to have ‘high titer’ anti-A/A,B: implications for transfusion
single-donor apheresis products reduces the risk that a recip- policy. Transfusion 2004; 44:805–808
ient is transfused with a high ABO agglutinin-titre, single-
donor product. Paolo Rebulla
Noemi Greppi
(2) The standard use of T-Sol reduces the absolute amount
Donatella Riccardi
of ABO agglutinins in the platelet products.
Fernanda Morelati
Based on the characteristics of our platelet products, Centro Trasfusionale e di Immunologia dei Trapianti
although we aim to use ABO-identical platelet transfusions IRCCS Ospedale Maggiore
whenever possible, this is not an absolute requirement. Of the Via Francesco Sforza 35
4146 products delivered in 2003, 425 pools were issued for 20122 Milano
inventory replacement to other institutions and 934 were Italy
prepared with buffy coats of different ABO/Rh groups. Of the E-mail: [email protected]
I88
International
nternational
3April 2005 Forum

remaining 2787, in which all buffy coats were of the same

ABO/Rh group and the recipient ABO group was registered
in our data management system, 237 (8·5%) were pools of
M. Uchikawa, N. H. Tsuno, K. Takahashi & K. Tadokoro
group O given to A, B or AB recipients.
Question 1
Question 2 We have no experience regarding haemolysis after trans-
We do not routinely take measures to prevent haemolysis, as fusion of apheresis platelets or after transfusion of pooled
outlined in answer 1. Nonetheless, we specifically ensure that concentrates.
an apheresis platelet product used for a neonate or for intra-
uterine transfusion is ABO identical. When this is not possible Question 2
and the platelet product shows plasma incompatibility with We make it a rule to transfuse ABO-compatible apheresis
the recipient, we concentrate the platelets by centrifugation platelet concentrates. However, in the case of human leucocyte
and replace the plasma with T-Sol immediately before infusion. antigen (HLA)- or human platelet antigen (HPA)-matched
platelets, if platelet concentrates of an appropriate ABO type
Question 3 are not available, we perform anti-A and/or anti-B titration
a) See above. of donor plasma.
b) Whenever we believe that the titre may be important
(i.e. very infrequently), it is determined by using 10 µl of Question 3
3% red blood cell (RBC) suspension and 40 µl of plasma or In our Blood Center, we determine the anti-A or anti-B titre
serum in the microcolumn agglutination technology by using the indirect antiglobulin test after incubation at
(BioVue; Ortho Diagnostic Systems, Raritan, NJ). We define 37 °C. Using this method, a titre of ≥ 1 : 512 is suggestive of
critical a titre in excess of 1 : 64 for immunoglobulin M a critical titre for anti-A or anti-B. If the anti-A or anti-B in
(IgM) and in excess of 1 : 256 for immunoglobulin G apheresis platelets exceeds the critical titre level, this is com-
(IgG), respectively [1]. We do not have a standard policy of municated to the clinicians. In this setting, some clinicians
permanent exclusion of donors from the donor panel for transfuse the platelet concentrate with a reduced volume of
transfusions to recipients whose red cells are incompatible plasma, whereas others prefer not to transfuse platelet
with anti-A/B in the platelet concentrate. We do not prefer concentrates with titres above the critical level.
donors with blood group A or B over donors with blood We try to administer platelet concentrates from group O to
group O. recipients of other ABO groups as rarely as possible. In the
c) See the answers to questions 1 and 2. past year, we supplied 2362 HLA-matched apheresis platelet
d) See the answers questions 1 and 2. concentrates, 567 of which were ABO-minor incompatible.
Only 4·9% (28/567) were supplied as group O to other ABO
Question 4 groups.
We aim to use ABO-compatible platelets for all recipients. In
2003, ABO major compatibility was achieved in 95·5% of Question 4
platelet transfusions. We do not routinely determine anti-A/ As it has been clearly shown that anti-A or anti-B in the
B titres in the recipients. recipient can considerably shorten the life span of incompat-
ible platelets, we use ABO-compatible platelets on a routine
Reference basis for all patients. However, when HLA- or HPA-matched
1 Josephson CD, Mullis NC, Van Demark C, Hillyer CD: Signifi- platelets are needed, the titres of anti-A or anti-B only in the
cant numbers of apheresis-derived group 0 platelets units donor plasma is measured. Moreover, in emergency situations,

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

 International Forum 215

in which ABO-incompatible platelets are transfused, the titres sion reaction was observed in a child of blood group AB after
of anti-A or anti-B are measured only in the donor plasma. transfusion of an apheresis PC of blood group O.
In 2003 the national haemovigilance foundation (TRIP)
Makoto Uchikawa
reported 186 reactions towards platelet transfusions and 697
Tokyo Blood Center Japanese Red Cross
reactions to red cell transfusions. They calculated a risk three
Nelson H. Tsuno times higher for transfusion reactions after platelet transfu-
Koki Takahashi sion than after red cell transfusions, but none was associated
Department of Transfusion Medicine and Immunohematology with haemolysis.
University of Tokyo
7-3-1 Hongo Question 2 and 3
Bunkyo-ku The blood bank and hospitals both aimed to circumvent (as
Tokyo 113-8655
much as possible) ABO-incompatible plasma with the PCs, but
Japan
the demand could not always be met, because mainly blood
E-mail: [email protected]
group O and A PCs are prepared. There is still no national
Kenji Tadokoro policy to determine the anti-A and/or anti-B titres in donors
National Headquarters Japanese Red Cross who contribute to buffy coat-derived or apheresis PCs. Two
I88
International
nternational
3April 2005 Forum

blood banks, providing PCs for four university hospitals,

prepare PC in additive solution.
V. M. J. Novotny & A. Brand One hospital mentioned that in all PCs, incompatible plasma
Question 1 was removed prior to transfusion by volume reduction (to
In order to respond to the topic of ABO-incompatibility of ≈ 20 ml for adults and < 10 ml for neonates). The blood
platelet transfusions on a national level, we questioned the bank prepared these concentrates on request. HLA/human
four Dutch blood bank divisions and the eight Dutch univer- platelet antigen (HPA)-matched platelet transfusions are more
sity hospitals for their policy and experience. often ABO-mismatched for minor as well as for major ABO
As shown in Table 1, in 2003 the blood banks delivered differences.
31 589 platelet concentrates (PCs) to the hospitals. The major- As contradicting data exists in the literature regarding
ity of these PCs were prepared from a pool of five random platelet loss during centrifugation to remove plasma, we asked
unit buffy coats and a minority [mainly for neonates and the aforementioned hospitals for clinical experience with
human leucocyte antigen (HLA)-matched PCs] were prepared volume-reduced PCs. They reported results on 533 matched PCs
as single-donor apheresis platelets. Of the eight university administered to 53 HLA (HPA) alloimmunized patients. In 69%
hospitals, six responded and five provided the total number (n = 368) of the transfusions, a minor ABO-incompatibility
of PCs transfused. In these five hospitals, 17 621 PCs had existed between the HLA-matched platelet donor and the
been transfused, representing 56% of all PCs prepared by the patient. All of these incompatible transfusions were volume-
blood banks. None of the interviewed hospitals reported reduced; the 1-h post-transfusion corrected count increment
haemolysis after platelet transfusion in 2003, although one (CCI) was similar to the increment of ABO-minor compatible
hospital remarked that in 2004 a severe haemolytic transfu- PCs in plasma, which were not volume-reduced. The 24-h
recovery tended to be slightly lower, a decrease that was
neither significant nor clinically relevant (Table 2).
Table 1 Results of the questionnaire of the Dutch blood banks and hospitals

Blood banks n=4 Table 2 Results from 53 alloimmunized patients: corrected count
increments (CCIs) of human leucocyte antigen/human platelet antigen
PC distributed (n) 31 589 (HLA/HPA)-matched platelet transfusions, either in plasma (minor ABO
Preventive measures compatible) or volume-reduced to 20 ml prior to transfusion (minor ABO-
Use of plasma-compatible PCs If possible incompatible)a
Exclude donors with high anti-A and/or B titres No
Use of platelet additive solution 50%a PC in plasma PC volume-reduced P-value
Reduction of the plasma volume by concentration 15%b
ABO antigen incompatibility measures No. of transfusions 165 368
ABO antigen-compatible PCs If possible CCI at 1 hr ± SD 13·4 ± 8·4 14·3 ± 9·3 0·31
Only when increased anti-A and/or B titres patient No No. of transfusions 102 184
CCI at 24 hrs ± SD 6·9 ± 6·8 5·4 ± 6·7 0·06
a
Two of the four blood banks provided platelets in additive solutions (PAS II).
b a
One blood bank, at the request of the regional university hospital. Personal communication, B. A. S. Tomson (Sanquin division Southwest).
PC, platelet concentrate. SD, standard deviation.

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

216 International Forum

Table 3 ABO antigen compatible vs. incompatible platelet transfusionsa B. G. Solheim

3IApril
88
International
nternational
2005 Forum

ABO identical and Major incompatible Question 1

minor incompatible n = 79 n = 14 P-value Haemolysis has been observed after the transfusion of
apheresis platelets, but not after the transfusion of pooled
1 h CCI 12·9 (± 6·7) 8·9 (± 4·9) 0·051 buffy coat platelets (which are resuspended in platelet addi-
24 h CCI 8·1 (± 6·7) 5·3 (± 4·7) 0·138 tive solution and ≈ 25% plasma).

CCI, corrected count increment.

a
Personal communication, J. L. Kerkhoffs (Sanquin Southwest). Question 2
In Norway we take measures to prevent haemolysis, caused
Question 4 by anti-A/B in platelet concentrates, in the recipients of
If possible, PCs are ABO-antigen matched, but in ≈ 10–20%, apheresis platelets.
the transfusion is ABO-antigen incompatible.
Although PCs are ABO antigen matched as much as possi- Question 3
ble, in particular for HLA/HPA-alloimmunized patients, pre- a) If available.
emptive ABO-antigen matching further impairs the number b) Saline agglutination is used to determine the titre of
of available matched donors. Neither for random PC nor for immunoglobulin M (IgM), and the indirect antiglobulin
HLA-matched PC are the anti-A and/or B titres of the recip- technique is used to determine the titre of immunoglobulin
ient routinely determined, but form part of a decision tree in G (IgG). We consider a titre of 1 : 250 to be critical. We do
the event of failure to reach a sufficient post-transfusion not permanently exclude donors with titres above the critical
increment; under these cicumstances, the titre of anti-A and/ level, and we do not prefer donors of blood group A or B over
or B is taken into account and if > 1 : 128 to 1 : 250, only donors of blood group O.
ABO-antigen matched PCs are further administered. c) We attempt to resolve the problem by reducing the
In a non-selected group of patients with AML and an un- volume of plasma.
known anti-A/B titre, who participate in a randomized, con- d) The plasma is replaced by platelet additive solution.
trolled study comparing PAS vs. plasma-stored PCs, a lower e) When apheresis blood group O platelets are considered
CCI was observed, in particular 1 h after transfusion. In this issued to an A/B blood group recipient, plasma from several
study, 411 PCs were transfused, of which 79 were ABO-minor platelet units is diluted 1 : 250 and tested for IgM and IgG.
incompatible and only 14 were major incompatible. Even with If no antibody is detected at this dilution, the platelet units
these small numbers the difference was significant (Table 3). are issued. If antibody is detected in a unit which has to be
In conclusion, overt haemolytic transfusion reactions upon transfused (i.e. HLA- or HPA-matched unit) the platelets are
ABO-minor incompatible platelet transfusions are rare. It should centrifuged and resuspended in platelet additive solution
be noted that serological effects, e.g. development of a positive shortly before transfusion.
direct antiglobulin test (DAT), was not recorded and subclin-
ical haemolysis is thus not excluded. Removal of incompatible
plasma can be safely carried out without obvious impairment Question 4
of platelet recovery. A policy to withhold ABO-antigen In Norway we use ABO-compatible platelets for all patients,
incompatible PCs should be maintained as ABO-antigen with the exception of blood group A2-platelets that are
incompatiblity impairs post-transfusion increments. considered clinically compatible also to patients who are not
blood group A. In selected patients HLA- or HPA-compatibility
V. M. J. Novotny
is preferred over ABO-compatibility.
Department of Blood Transfusion and Transplantation
Immunology
Bjarte G. Solheim
Radboud University Medical Centre
Rikshospitalet University Hospital
PO Box 9101
University of Oslo
6500 HB Nijmegen
NO-0027 Oslo
the Netherlands
Norway
E-mail: [email protected]
E-mail: [email protected]
3IApril
88
International
nternational
2005 Forum

A. Brand
Sanquin Blood Bank South West B. Zupanska
PO Box 2184
2301 CD Leiden Question 1
the Netherlands I have not seen cases of haemolysis after transfusion of
E-mail: [email protected] platelets.

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

 International Forum 217

Question 2 For pooled PCs, before using a platelet additive solution,

As a rule, we transfuse platelets from ABO-identical donors. we applied the same strategy as in apheresis products.
Regarding the patients who require human leucocyte antigen Currently, with leucoreduced pooled buffy coat-derived PC
(HLA)- or human platelet antigen (HPA)-matched platelets, in additive solution with a plasma carryover of ≈ 30%, we no
see the answer to question 3. longer take the plasma ABO group into account.

Question 3 Question 4
We do not determine the titre of anti-A/B in donors, because We try to transfuse ABO group identical PC whenever possible
we transfuse platelets from ABO-identical donors. for all patients. When this goal cannot be accomplished, our
For patients refractory to platelet transfusions and with priority is to avoid or significantly reduce the infusion of
HLA antibodies, we select compatible platelets (by cross- ABO-incompatible plasma to the recipient while maintaining
matching with lymphocytes and sometimes also with platelets) compatibility between ABH platelet antigens and recipient
from ABO-identical donors. Very rarely do we select HLA- isohaemagglutinins [2].
matched/partly matched platelets or platelets without the
antigen against which the antibody is suspected to be or References
really is directed. We try to find an ABO-compatible donor. 1 Moroff G, Friedman A, Robkin-Kline L, Gautier G, Luban NL:
If we have to give ABO-incompatible platelets, we remove/ Reduction of the volume of stored platelet concentrates for use
reduce the volume of the donor plasma and replace it with in neonatal patients. Transfusion 1984; 24:144–146
2 Lozano M, Cid J: The clinical implications of platelet transfu-
AB plasma or with plasma of a donor which does not contain
sions associated with ABO or Rh(D) incompatibility. Transfus
antibodies against the recipient’s A or B antigens. We do not
Med Rev 2003; 17:57–68
replace the plasma by platelet additive solution because this
has not yet been registered in our country. Miguel Lozano
To date we have transfused HPA-matched platelets, found Roberto Mazzara
in our registry, to five patients with antibodies to HPA, refractory Department of Hemotherapy and Hemostasis
to random platelets. All of these patients were ABO compatible. Agustí Pi i Sunyer Biomedical Research Institute (IDIBAPS)
Hospital Clínic
Question 4 University of Barcelona
As mentioned above, we use ABO-compatible platelets for Villarroel 170
almost all patients. If we have to give incompatible platelets, 08036 Barcelona
Spain
we choose O-group platelets resuspended in AB group plasma.
E-mail: [email protected]
Barbara Zupanska
Joan Cid (present address)
Institute of Haematology and Blood Transfusion
Blood Transfusion Center and Tissue Bank
5 Chocimska Str. 00957 Warsaw
Barcelona
E-mail: [email protected]
3IApril
88
International
nternational
2005 Forum Spain
3I88
International
nternational
April 2005 Forum

M. Lozano, J. Cid & R. Mazzara F. Knutson & R. Norda

Question 1 Question 1
No, we have not seen cases of haemolysis after transfusion No incidents of haemolysis after transfusion of apheresis
of platelets. platelets or of pooled concentrates have been reported to the
Swedish haemovigilance system.
Question 2
Some time ago in our centre we implemented preventive meas- Question 2
ures to avoid haemolysis caused by ABO-mismatched platelet Measures are taken to prevent haemolysis, caused by anti-A/
concentrate (PC) transfusion, a fact that could explain why B in platelet concentrates, in the recipients of apheresis plate-
we have not seen any cases of haemolysis in recent years. lets. However, if human leucocyte antigen (HLA)- or human
platelet antigen (HPA)-matched platelets are required, that is
Question 3 our first concern.
To prevent haemolysis in recipients caused by a minor ABO-
mismatched apheresis product, for a volume higher than 150 ml Question 3
we reduce the plasma volume by centrifugation [1] immedi- a) No, we do not transfuse platelets only from ABO-identical
ately before transfusion, yielding a final volume of ≈ 90 ml. donors.

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

218 International Forum

b) All our apheresis platelet donors with blood type O are platelet concentrates (SDPC), 20% pooled buffy coat platelet
checked for anti-A/B and for HLA antibodies. Titration of concentrates (BCPC) to 61 hospitals and transfusion services.
anti-A/B immunoglobulin M (IgM) and immunoglobulin G We are not aware of any case of hemolysis due to platelet
(IgG) is performed by using the tube-technique [1]. Critical transfusion across minor ABO incompatibility. However, we
titres are believed to be 1 : 100 for IgM and 1 : 400 for IgG. are not directly involved in the administration of products to
We do not permanently exclude donors with titres above the patients and therefore might miss clinical outcome information.
citical level. A low titre of both IgG and IgM means accept-
ance as a universal donor, whereas a high titre of one or both Question 2
means acceptance as a donor for a recipient of the same ABO Yes. In collaboration with the clinicians, we select the PCs for
type. We have no preference with respect to ABO-type. The transfusion according to the ABO group of recipient. How-
majority of our donors are low-titre donors with blood type ever, for logistical reasons, we provide only PCs of group A
O and the rest are of blood type A. and O. Therefore, ABO-mismatched transfusions of PC will
c) No, we don’t resolve the problem by reducing the occur in many cases.
volume of plasma.
d) We are planning to replace the plasma with an additive Question 3
solution for quality reasons. Each donor of an SDPC is tested for the presence of anti-A
e) No other measures are used or planned. and anti-B hemolysins. If the titre is > 1 : 16, the product will
be labelled and used for ABO-matched receipients only.
Question 4 For BCPC the plasma will be replaced by platelet additive
We use ABO-compatible platelets for all patients. solution. Therefore, these products are not tested for hemolysins.
Technically, the hemolysins are determined by using test
Reference cells of type A1 and B. We use a microtiter plate format (96-
1 Anon: Serological Techniques. Houston, TX, USA, Gamma well plate) and measure the haemolytic reaction photometrically.
Biologicals, Inc., Revised November 1996 There is no policy to defer hemolysin-positive donors from
apheresis donation. However, for every apheresis donation
Folke Knutson the donor will be retested for hemolysins and the product will
Clinical Immunology and Transfusion Medicine
be labelled accordingly.
University Hospital
SE-751 85 Uppsala
Sweden
Question 4
E-mail: [email protected] ABO-compatible PCs will be delivered according to the
transfusionist’s request. However, we believe that ABO-
Rut Norda compatible PCs are mainly indicated in cases needing
Clinical Immunology and Transfusion Medicine prophylactic or chronic platelet substitution or carrying
University Hospital antibodies to human leucocyte antigens (anti-HLA) or to
SE-751 85 Uppsala human platelet antigens (anti-HPA). For therapeutic use,
Sweden ABO-minor-mismatched PCs without significant amount
E-mail: [email protected] of isohemolysins seem to work well.
I88
International
nternational
3April 2005 Forum

Beat M. Frey
B. M. Frey Zürich Blood Transfusion Service
Hirschengraben 60
The Zürich Blood Transfusion Service of the Swiss Red Cross is
CH-8001 Zürich
the largest Regional Transfusion Service of the country regard- Switzerland
ing production and distribution of platelet concentrates (PC). E-mail: [email protected]
Annually, we produce and deliver more than 5500 units to 3April 2005
88

about 65 hospitals and transfusion services. Since years, we International Forum

provide about 80% single donor apheresis units (SDPC) and

S. MacLennan
20% pooled buffy coat units (BCPC). As a responsible Medical
Director and CEO of the Zürich Blood Transfusion Service, I am Question 1
delighted to be able to contribute to this International Forum. The 2003 Serious Hazards of Transfusion (SHOT) survey
reported two incidents of significant haemolysis caused by
Question 1 apheresis platelets, both of which were from hospitals served
Annually, the Zürich Blood Transfusion Service delivers by the National Blood Service (NBS) [1]. One incident
about 4500 units of platelet concentrates (80% single donor concerned a 3-month-old infant postcardiac surgery, who

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

 International Forum 219

was given multiple transfusions of O Rhesus D-negative plate- This technique nationally has been found to give an over-
lets owing to a shortage of A Rhesus D-negative neonatal all reactive rate of 10%.
platelets. Although the donations were tested and found to A further initiative with regard to high-titre testing is
be negative for high-titre anti-A/B, it is possible that the being taken forward by the Standing Advisory Committee on
haemolysis occurred as a result of repeated transfusion. The Immunohaematology for the UK Blood Transfusion Services.
other incident involved a 31-year-old man, who was trans- It is developing two standard reagents (positive control and
fused with Group O platelets. Retrospective testing of the negative control) for use by the UK Blood Services for assess-
donor demonstrated high titres of both immunoglobulin G ing the cut-off of their testing methods. The controls are
(IgG) and immunoglobulin M (IgM) anti-A. designed to detect anti-A/B, with a cut-off of 1 : 128, by
In the previous 5 years of SHOT reports, a total of six using the manual tube technique.
haemolytic reactions to platelet transfusion were reported [2]. Donors found to be HT positive are not specifically excluded
from the panel, but their donations are not selected for trans-
Question 2 fusion to patients of incompatible groups. In addition, these
Yes, the NBS has a national protocol for testing all donations (not donations are not used for manufacturing any neonatal
just apheresis platelets) routinely for the presence of anti-A/B. components.
The majority of platelet apheresis donors in the NBS are
Question 3 group O or group A. Recruitment has not been targeted spe-
a) The NBS has drawn up a clinical policy entitled ‘ABO cifically for A and B donors in preference to those of group O.
and Rh D compatibility in relation to platelet transfusion’ Hospitals are advised that group O platelets with high titres of
(available on the NBS website: www.blood.co.uk/hospitals/ anti-A or anti-B should be transfused only to group O recipients.
guidelines/index.hm) [3]. This recommends transfusing ABO- c) Plasma reduction of platelet components is not currently
identical platelets whenever possible, but recognizes that this employed.
is not always possible owing to additional special requirements, d) It is planned to implement the use of platelet additive
e.g. gamma-irradiated, cytomegalovirus (CMV)-seronegative, solution routinely during the next year. This may offer
or shortages of a group. If a group A platelet is requested and advantages for both TRALI and variant Creutzfeldt–Jacob
not available, then a group B platelet is offered in preference disease (vCJD) risk reduction in addition to reducing the
to group O, and vice versa. If incompatible platelets are being anti-A/B content.
transfused, then they should be labelled ‘high-titre negative’
(HT neg; see below). Question 4
b) A national procedure for high-titre testing has been As outlined above, ABO-compatible platelets are used whenever
in place in the NBS for several years. Until very recently the possible. We do not routinely measure the titre of anti-A/B in
method used was an inhibition method, which used AB patients, including bone marrow transplant recipients.
substance to neutralize IgM in most samples, and those that
then still agglutinated cells were considered ‘high titre’. This References
test identified ≈ 5–10% of donations as ‘high titre’ and these 1 Stainsby D, Cohen H, Jones H, Knowles S, Milkins C, Chapman C,
were directed for use in patients of the compatible group Gibson B, Davison K, Norfolk DR, Taylor C, Revill J, Asher D,
only. This method has now been withdrawn because AB Atterbury CLJ, Gray A: Serious Hazards of Transfusion (SHOT)
substance as a reagent is no longer available, and a new Annual Report 2003. Manchester, Serious Hazards of Trans-
method has been implemented. fusion Office, 2004
2 Asher D, Atterbury CLJ, Chapman C, Cohen H, Jones H,
The new method is performed on the Olympus PK7200, on
Love EM, Norfolk DR, Revill J, Soldan K, Todd A, Williamson LM:
all donations, in parallel with routine donation testing. All
Serious Hazards of Transfusion (SHOT) Annual Report
steps are fully automated and tested on the Olympus micro-
2000–2001. Manchester, Serious Hazards of Transfusion
plates. Fifteen microlitres of a 1 : 20 dilution of donor plasma Office, 2002
in phosphate-buffered saline is added to 25 µl of A2B cells. 3 Murphy MF: ABO and RhD Compatibility in Relation to Platelet
The detection of any agglutinated red cells is a positive result. Transfusion. NBS, Oxford Centre, NBS Clinical Policies
A negative result is recorded when there is no agglutination, Group, 2001
and these donations are labelled as ‘HT neg’. Positive donations
are not specifically labelled. This means that if a hospital
Sheila MacLennan
blood bank knows that ABO-incompatible platelets are being National Blood Service
given, it can select a donation labelled as ‘HT neg’. This Bridle Path
applies to pooled platelets (all of the contributing donations Leeds, LS15 7TW
must test HT neg to have the final component labelled as UK
such) as well as apheresis platelets. E-mail: [email protected]
I88
International
nternational
3April 2005 Forum

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

220 International Forum

J. M. Heal & N. Blumberg saline-resuspended, machine-washed O platelets preferen-

tially. There are minimal quantities of incompatible cells,
Question 1
soluble antigen and antibody in this transfused component.
We have not seen clinically evident cases of haemolysis
We also sometimes decrease the number of transfused whole-
after transfusion of incompatible platelets in recent years, as
blood platelet concentrates from a routine pool of five units
> 90% of our platelet transfusions are ABO identical and we
to three or four, so that only ABO-identical transfusions are
only rarely administer group O platelets to non-O recipients,
administered.
for the reasons discussed below. Prior to 1990, when we
We do not use titres to select mismatched platelets
routinely administered ABO-mismatched platelets, we fre-
because we believe that this method is unreliable at pre-
quently observed patients with positive direct antiglobulin
dicting biological and clinical effects in the recipient. We
tests, increased red cell transfusion needs and other evidence
are concerned that employing titres of incompatible ABO
of haemolysis.
antibodies to select ‘safe’ donors may represent treating
ourselves rather than effectively treating the patient,
Question 2 given the lack of data supporting the predictive value of
For unavoidable ABO-mismatched platelet transfusions, as titres.
in patients receiving human leucocyte antigen (HLA)- or
human platelet antigen (HPA)-matched platelets, we wash Question 4
before transfusing to remove the incompatible supernatant We are of the opinion that the use of the term ‘compatible’,
plasma antigen and antibody. While some might consider derived from red cell transfusions that are relatively plasma
this plasma reduction to be unnecessary, we are increasingly poor, should not be applied to platelet transfusions that
convinced that the administration of large quantities of contain an order of magnitude greater amounts of plasma-
incompatible anti-A and anti-B, or soluble A and B antigens, soluble antigen and antibody. Platelet transfusions are more
may have deleterious effects on many patients that are not analogous to whole-blood transfusions in this regard, and in
obvious [1]. the modern era, ABO non-identical whole blood would rarely
Some clinical events that we hypothesize are caused by the or never be considered suitable for transfusion. Unless
transfusion of ABO-incompatible plasma and cells are not as plasma reduced, we believe that platelet transfusions should
easily attributable to the transfusion as post-transfusion ideally be ABO identical.
haemolysis. For example, the transfusion of ABO-mismatched In our centre, thrombocytopenic patients with haemato-
platelets increases the risk of HLA alloimmunization and plate- logical diseases are transfused solely with ABO-identical or
let refractoriness in the two small randomized trials address- washed platelet antigen and antibody-identical platelet
ing this issue [2,3]. Transfusion of incompatible plasma leads concentrates. We administer a reduced dose of whole blood-
to large quantities of circulating high-molecular-weight, derived platelet concentrates (e.g. a pool of four rather
long-lived immune complexes. These immune complexes fix than of five), rather than give incompatible plasma or plate-
complement, bind to platelets, leading to phagocytosis by lets to the patient. We also transfuse solely ABO-identical
monocytes, and carry unknown potential for morbidity [4]. platelets to patients with ventricular assist devices who are
Such morbidity speculatively could include interference with awaiting cardiac transplantation, because their platelet
anti-leukemia cellular immunity [5] or a pro-inflammatory transfusion needs are considerable over a period of days to
predisposition to multi-organ failure and death in cardiac weeks, somewhat analogous to patients with haematological
surgery [6]. diseases.
The assumption that the sole biological and clinical effect For all other patients we attempt to give only ABO-identical
of transfused anti-A and anti-B is to cause destruction of platelets, but occasionally no such platelets are available.
circulating red cells seems improbable to us. ABH antigens Under these circumstances we will transfuse unwashed
are present in the recipient in soluble form, on endothelial ABO-mismatched platelets. This happens most commonly
cells, white cells and virtually every other cell in the body. in emergency transfusions for massive haemorrhage in
Transfusion of incompatible soluble A and B antigens may trauma or liver transplantation, where there is no time for
likewise not be benign. washing or plasma reduction by centrifugation. The inci-
dence of this is fewer than 5–10% of transfusions in our
Question 3 institution.
We attempt to transfuse only ABO-identical platelets, partic-
ularly to patients receiving repeated platelet transfusions. References
When non-ABO identical platelets must be administered 1 Heal JM, Blumberg N: The second century of ABO: and now
owing to shortages or changing ABO blood groups in allo- for something completely different. Transfusion 1999;
geneic stem cell transplant recipients, we routinely transfuse 39:1155–1159

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221

 International Forum 221

2 Carr R, Hutton JL, Jenkins JA, Lucas GF, Amphlett NW: Joanna M. Heal
Transfusion of ABO-mismatched platelets leads to early Associate Clinical Professor of Medicine
platelet refractoriness. Br J Haematol 1990; 75:408–413 Hematology-Oncology Unit
3 Heal JM, Rowe JM, McMican A, Masel D, Finke C, Blumberg N: Department of Medicine
The role of ABO matching in platelet transfusion. Eur J University of Rochester
Haematol 1993; 50:110–117 Rochester
4 Heal JM, Masel D, Rowe JM, Blumberg N: Circulating NY 14642
immune complexes involving the ABO system after platelet USA
transfusion. Br J Haematol 1993; 85:566–572 E-mail: [email protected]
5 Heal JM, Kenmotsu N, Rowe JM, Blumberg N: A possible
survival advantage in adults with acute leukemia receiving Neil Blumberg
ABO-identical platelet transfusions. Am J Hematol 1994; Transfusion Medicine
45:189–190 University of Rochester
6 Blumberg N, Heal JM, Hicks GL Jr, Risher WH: Association of Box 608
ABO-mismatched platelet transfusions with morbidity and Rochester
mortality in cardiac surgery. Transfusion 2001; 41:790–793 NY 14642
USA
E-mail: [email protected]

© 2005 Blackwell Publishing Ltd. Vox Sanguinis (2005) 88, 207–221