& The rapid development in bloodcomponent technologies has lead toproduction of various types of plateletconcentrates which are highly hetero-genous in terms of cellular content,subpopulation of platlet and leucocytesand in vitro characteristics and storagestability. This emphasizes the need for astringent in-process validation and sta-tistical process control to ensure thatoptimal process ef®ciency and plateletfunctional integrity are continuouslyattained and maintained throughout the5 days shelf life. Several unresolvedissues such as development of cytokinesand microparticles, some with immu-nomodulatory effects; fears of bacterialcontamination of leucodepleted plate-lets, in particular in storage media; lackof progress in reduction of someunquanti®ed risk of all types of transfu-sion transmitted infection (TTI), stillremain the focus of current interest.While some processing results recentlysucceeded in killing both viruses andbacteria through a recently developedviral inactivation and ®ltration process,nevertheless much still remains to beelucidated in respect to long termeffects of such treatments. Apart fromdevelopment of purer products to reducetransfusion reactions, some compli-mentory alternatives to platelet trans-fusion are becoming available, which in

the near future, may in¯uence thedemand for platelet support. # 1997Elsevier Science Ltd &

Several variants of the three traditionalmethods of platelet production on con-centrates derived from either plateletrich plasma (PRP-PC), pooled buffy coat(BC-PC) and haemapheresis technology(H-PC) are currently being used to meetplatelet demand. Today in Europe thelong-standing method of PRP-PC hasbeen replaced by pooled BC-PC, whichuses a semi-automated separation ofwhole blood in a specially designed bot-tom and top pack. This approach helpedto a great extent in the standardisationprogramme. Moreover with the ®netuning of the second centrifugation stepit is possible to reproduce with consis-tency, a ``standard adult platelet dose''(S-APD), containing on average 300�109

PLT/pack, 30±40% of which are leuco-depleted (WBC �5�106/pack). As forhemapheresis products there is a clearmove towards the production doubledose leucodepletal PC, with a plateletcontent of above 2.4� 1011 per pack.This manuscript deals with a personalview of current opinions on bothlaboratory and clinical aspects of plate-let therapy, in particular, the future ofleucodepletion, platelet storage mediaand potential clinical signi®cance ofcytokines and other biological responsemodi®ers (BRM) are discussed together

Transfus. Sci. Vol. 18, No. 3, pp. 345±351, 1997# 1997 Elsevier Science Ltd. All rights reserved

Pergamon Printed in Great BritainPII: S0955-3886(97)00029-5

345

Platelet Therapy: Current Opinionson Laboratory and Clinical

AspectsJerard Seghatchian, PhD*

*National Blood ServiceÐNorth London, CollindaleAvenue, London NW9 5BG, U.K.

with the in¯uence of the alternatives toplatelet support.

PRODUCT HETEROGENEITY

In terms of the laboratory characteristicproperties a much larger subpopulationof platelets are present in the BC-PCand H-PC as compared to PRP-PCwhich often contains large amounts ofaggregates (clumps). However the pH ofBC-PC is usually about 0.3pH unitslower than that for other types of PC ofa similar age. This contributes to theabsence of aggregates in BC-PC com-pared to the standard H-PC (i.e. Hae-monetics PCS/MCS) which uses thesame principle. However some BC-PCdo not store well up to 5 days, requiringpacks with a suf®ciently larger surfacearea for optimal clinical effectiveness.

Several modi®ed haemapheresisprocedures have been used to ``top up''the special requirements. Most of thecurrently available haemaphersismachines can provide double dose, butrequire leucocyte ®ltration and otherssuch as the COBE Spectra-LRS uses anew harness design allowing it to pro-duce a double dose PC which is with99.9% consistency leucodepleted.

While once platelet products wereonly compared on the basis of cost ofcollection and cellular content todayopinions revolve around the shorteningof storage period for the fear of bacterialgrowth and generation of cytokines; usebefore date policy the prereleased assess-ment of functional/tructural integrity toreduce the frequency of transfusionreactions are becoming important issuesfor clinical decision making. Interest-ingly some cardiac surgeons are reluctantto use standard PC for neonates becauseof marked adverse effects possibly dueto generation of cytokines during sto-rage. In response, a concentrate in 40±60 ml can be effectively produced by theCOBE-Spectra LRS with good storagestability for 3 days. Haemapheresis pro-vides a unique approach for specialiseddesigner product upon demand.

UNRESOLVED ISSUES

From the clinical stand point the effec-tiveness of platelet therapy is based onful®lment of three important criteria:

(i) There must be suf®cient num-bers of platelets to increase thecirculatory platelet countabove a threshold at whichspontaneous bleeding does notoccur. In the light of the het-erogeneity in platelet prepara-tions, it is therefore essentialto reappraise the real platelettransfusion trigger point.

(ii) The transfused platelet mustapproach the native (discoidstate) in the circulation and beable to circulate for a period of7±10 days, and to retain thecapacity for correcting hae-mostasis. Thus it is essentialthat the morphological, struc-tural and functional integrityof platelets during 5 days sto-rage are well maintained. Thisis of particular relevance toprophylactic use where bothsurvival time and hemostaticcapacity are of equal impor-tance and clearly differentiatenative platelets from 4�Cstored- PC, lyophilised plate-lets, membrane vesicles andsynthetic liposomes with pla-telet glycoprotein complexeswhich are all able to correctbleeding time, but having dif-ferent survival times.

(iii) A product which is both aggre-gate-free and containing lessmicrovesicles or soluble anti-gens (HLA), cellular metabo-lites, hence with reducedcapacity of ``non-immuno-genic'' transfusion reactions. Inthis respect any additional pro-cessing has some deleteriouseffect on platelet morphologi-cal structural and functionalintegrities, often leading toformation of microvesicles.

346 Transfus. Sci. Vol. 18, No. 3

While it is still unclear to whatextent variable degrees of platelet acti-vation and storage lesion, may in¯uenceboth the rate of clearance of plateletsfrom the circulation (i.e. the overallclinical outcome), nevertheless theclinicians are focusing on the issues ofplatelet dose in particular for conveni-ence of prophylactic, donor exposure pertransfusion and the management ofalloimmunised/non-alloimmunisedpatients. These clearly in¯uence thechoice of specialised products despitethe fact that the survival of such pro-ducts produced by different Haema-pheresis machines may not be identicalin all types of patients.

The recent development in the viralinactivated process and the use of plate-let storage media and leucocyte ®lters ofvarious principles, matrix and design,(some leading to considerable loss of largeplatelets and activation of haemostaticcomponents) makes it essential to assessthe activation, aggregation and micro-vesiculation of platelet preparation,hence requiring further in depth analyses.

SELECTION OF PLATELETS FORALLOIMMUNISED RECIPIENTS

Clinical response to platelet transfusionis measured by the increment in theplatelet count over time. A successfulCCI should be greater than 7500 within10±60 min and greater than 4500 ifmeasured 18±24 h after transfusion.Often the 1 h CCI is used for indirectmeasurement to document alloimmu-nisation vs non-immune causes of pla-telet transfusion failure.

However it should be borne in mindthat many factors are involved in plate-let refractoriness, in¯uencing the 1-hincrement. The fact that the speci®cityof HLA antibodies, in multiple-trans-fused individual, is generally againstpublic epitopes it supports the notionthat matching for the public antigens isimportant. In general the ability to ®ndcompatible donors is very low if apatient has an antibody to an HLA pub-

lic speci®city. Therefore HLA-matchingand pre-transfusion compatibility testing(i.e. testing the transfusion recipient'ssera against prospective platelet donors)is required for selecting platelet foralloimmunised patients.

THE FUTURE OF LEUCODEPLETION

The leucocyte burden per unit of wholeblood is, on average, 4�109. This is sub-stantially reduced in platelets derivedfrom buffy coats, containing �0.04�109

WBC/pack. While the generation ofcytokines is dramatically reduced insuch products, the critical immunologi-cal leucocyte load, de®ned as the con-centration of leucocytes necessary tocause sensitisation, is in a previouslynon-sensitised individual less than onemillion (�1�106 unit). Thus leucode-pletion on demand might stay with usfor some time to come.

Leucoreduction can be carried outnear the patient's bedside (post storage)or in blood centres on demand, usingeither sterile connecting devices or a®lter as an integral part of the pack sys-tem. The bedside processing has inher-ently more variability and is lesssubjected to stringent QC and GMPstandards than are in place at bloodcentres. Moreover prestorage ®ltration iscurrently favoured as in¯ammatorycytokines production is substantiallyreduced by removing leucocytes at anearly stage of storage.

The removal of leucocytes in theearly stages of storage (i.e. after 6±8 h)reduces the potential risk of viral/bac-terial contamination due to WBCbreakdown and leads to overall qualityimprovement in terms of released reac-tion induced changes and immuno-modulation. The use of a ®lter as anintegral part of the pack system is thepreferred strategy as it can potentiallyreduce the perceived risk of bacterialcontamination which may arise subse-quent to failure of sterile connectingdevices and the loss of precious compo-nents due to leakage.

Platelet Therapy: Current Opinions on Laboratory and Clinical Aspects 347

Allogeneic leucocytes also have acritical role in the induction of transfu-sion associated immunosuppression astransfusion-induced infection complica-tions is less frequent in patients receiv-ing leucocyte reduced products. Naturalkiller (NK) cell function is also signi®-cantly impaired, with allogenic leuco-cytes. NK cells are known to kill bothgram positive and negative bacteria bydirect contact and secretion of a lyticsubstance though there is a substantialvariability in antibacterial activity andNK-inhibitory activity of both donorsand recipients. This positive bonus of®ltration should nevertheless put inbalance with some negative points asevidence is accumulating that the pas-sage of platelet concentrates throughsome negatively charged ®lters canenhance by two to three times the Kal-likrien/Kinin-like activity and therelease of some microparticles due shearstress. These might be responsible for®ltration-induced anaphylactoid reac-tions. Obviously more work is stillrequired to de®ne the best practice inrespect to leucoreduced preparation,indications for the use and necessarylevels of leucocyte removal in order tobalance the cost bene®ts of this proce-dure. Moreover the new generation ofapheresis machines can produce with-out ®ltration leuodepleted productsmeeting this criteria though they mightnot remove the same sub populations ofleucocytes. Obviously for quality assur-ance purposes (process ef®ciency) sys-tems, the process control of leuco-reduced blood components is essentialto prospectively predict that the processbeing used will consistently lead to PCthat meet all criteria set for effectiveand optimal in vitro and in vivo functionas well as minimal frequency of poten-tial transfusion reactions.

THE FUTURE OF PLATELETADDITIVE SOLUTIONS

Several storage media have been intro-duced in order to improve PC quality

during storage and hence theirclinical effectiveness. The practicaladvantage of storage media includes: (i)the reduction of the platelet storagelesion caused by the rapid accumulationof lactate in particular when long termstored BC are used for BC-PCpreparation, (ii) the removal of someactivated plasma components such asthrombin, plasmin, complement andKallikrein/Kinin generation and (iii)the partial removal of cell-releasedenzymes and microvesicles, in particu-lar, those containing soluble HLANP.Currently storage media containingsodium acetate are the preferred choice,however the oxidation of acetaterequires: higher oxygen supply and theuse of high oxygen-permeable con-tainers.

Recently, a new storage media PAS2 became licensed in the U.K. Preli-minary validation experiments showedthat while platelet in this storage mediaretained their ligand binding capacityfor vWF, the binding capacity for ®bri-nogen is substantially decreased and asigni®cant difference in various proper-ties occurred when platelets in plasmaand in storage media were ®ltered toremove leucocytesÐthough the overallrate of deterioration appears to be simi-lar. Since the capacity of platelets tobind ligands like ®brinogen and vonWillebrand factor is most likely to be anindication of their capacity to adhereand aggregate at the site of injury fur-ther comparative studies are needed torelate the in vitro function of plateletsto their in vivo function. The character-istic ®lterability of concentrates inplasma vs storage media are also quitedifferent, requiring further investiga-tion. Clearly more work is needed totranslate the in vitro changes to clinicalef®cacy and the frequency of transfu-sion reactions, where the expression ofactivation speci®c markers might proveto be essential in differentiating variousproducts, when products are fresh orwhen have undergone some injury sub-sequent to various processing andstorage.

348 Transfus. Sci. Vol. 18, No. 3

CYTOKINES AND OTHERBIOLOGICAL RESPONSE MODIFIERS

The true biological signi®cance of cyto-kine accumulation in platelets duringstorage has not been well established.However there is a direct relationshipbetween the leucocyte content,increased levels of some cytokines andtransfusion reactions. Moreover it is notknow to what degree different subsets ofleucocytes may modulate and in¯uenceplatelet activation and cellular lesionduring storage.

The assessment of the leucocytesubsets may be a better predictor ofclinical outcomes such as alloimmuni-sation than the total number of leuco-cytes. The in¯uence of leucocytederived enzymes on changes in theexpression at activation markers onplatelet membrane has been previouslyindicated but the inherent transientnature of these markers has made theinterpretation of data rather unreliable.Thus it is becoming important to com-pare the expression of activation mar-kers on the platelet membrane beforeand after stimulation with weak andstrong agonists and relate them to theformation of microvesicles in thesupernatant. Certain additional proces-sing such as resuspension in storagemedia and/or UV/gamma-irradiationmay lead to shedding of microparticlesand release of proaggregatory substancesand/or activation of haemostatic com-ponents.

Some anaphylatoxin such as Kinin/Kallikrein and complement are abun-dantly found in platelet concentrates,after ®ltration. The effect of such com-ponents in concentrates on induction ofimmunosuppression and impaired car-diac function remains to be fully eluci-dated.

CONTINUAL QUALITYIMPROVEMENT

Improvement in quality will result inimproved productivity. To achieve this

there must be a system which canpredict accurately the collection/pro-cessing and storage-induced injury, onevidence based results. Amongst majorchallenges that contributing to contin-ual quality improvement in BTS are: (i)the application of statistical processcontrol (SPC) to improve the quality ofperformance in the whole blood trans-fusion organisation. Without SPC eventhe useful data may be interpreted in amisleading way leading to inaccurateadjustment to a satisfactory process, (ii)the application of ®shbone diagram,based on six important quality parame-ters (manpower, machine, material,method, matrix, milieu) in a targetorientated approach to preampt the fail-ure and to help in quality improvement.It is only through evidence based cause/effect relationship, continual qualityimprovement is achievable, (iii) appli-cation of Organisation-wide ClearInformation Design so called ``OrClD'',which is based on a standard for writingclearly and validating SOPs. This coreapproach which has been sponsored andendorsed by the NBA Executive facili-tates the process of sharing SOPs andpromote best practice across the servicenationally. As for any standard, OrClDstandard will be subject to continualreview and development to ensurethat it meets the organisation needs.Obviously working together, for estab-lishing and implementing the best prac-tice, does work better when qualityimprovement is concerned.

ALTERNATIVE THERAPY ANDCOMPLEMENTARY STRATEGIES

The future strategies should take intoaccount both the current and newdevelopments. Currently several otheralternatives such as freeze-dried plate-lets, IPM, procoagulant vesicle, univer-sal platelet for all blood groups are invarious stages of development and/orclinical trials. The implementation ofviral inactivated products has alreadybeen on the way forward for some time

Platelet Therapy: Current Opinions on Laboratory and Clinical Aspects 349

so is the ``Designer products'' made forselected groups of patients. New gen-eration methods become available fordifferentiating PC on the basis of plate-let functional integrity, activationstates and microvesiculation. Despiteenormous improvements in standardsthe competitive position on the marketnot yet fully taken up, but the intro-duction of some complementary ther-apy may change the continuous upwarddemand in platelet support. For examplenew development currently in progresson peripheral blood stem cell transplantprotocol to reduce the duration ofthrombocytopenia and transfusiondependent. The use of preplanned cyto-kine therapy such as thrombopoietin(Tpo) to stimulate marrow production ofplatelets has been introduced. It appearsthat when Tpo is bound to its speci®cligand to cmpL (a receptor on megakar-yocytes precursors), proliferation andmaturation to megakaryocytes rich inplatelet buds occurs. There are severalopportunities for using Tpo at the timeof initial cytotoxic therapy to inducethe remission in autologous transplants.The harvested cells could then expandmegakaryocyte precursors in the har-vest returned to the patient. Other pos-sibilities to include myelogenousmalignanciesÐalthough the possibleexpression of cMPL by the malignantcells requires caution.

At the present leukaemias andlymphomas constitute less than 10% ofall malignant disease in the U.K. andintensive myeloablative cytotoxic ther-apy is becoming a routine practice for anincreasing number of solid tumours.The transfusionists are considering sev-eral new complimentory developmentssuch as the use of TPO on megakaryo-cyte growth and development factor(MGDF) in their future plans.

SUMMARY

Continual quality improvement inblood components has led to the devel-

opment of tailor-made platelet concen-trates. New blood pack design andautomated processing and new develop-ment in the haemapheresis harness forleucodepletion, and new storage solu-tions have been recently introduced toimprove the clinical effectiveness andto reduce the frequency of transfusionreactions. This in turn led to productswhich are heterogeneous in respect toplatelet content and function. Onceproducts were only compared on thebasis of cost of collection, ``issue bydate/use before date'' policy. Today,however, the clinician prefers apheresisPC as it functions better than thoseprepared from units of whole blood Thisis supported by the fact that plateletincrements (1 h) after transfusion is bet-ter with fresh Haemapheresis PC andless than 3 days stored PC is superior tothe older PC. Stored PC also show two-fold higher frequency of transfusionreactions. There is a continuous con-cern about bacterial contamination andthe viral inactivation process shouldtake account of some unquanti®ed riskof all types of TTI. While the new inac-tivation process has given promisingresults in killing viruses and bacteriawithout grossly damaging platelets, insuspension media containing 30%plasma, the toxicity however, stillrequires further investigation beforesuch approaches can be applied to clini-cal practice. The use of Tpo or MGFDfor in vitro expression have also shownsome processing results. New genera-tion of techniques are currently in placewhich can de®ne quality on the basis ofmorphological/functional integrity andmicrovesiculation and their relation-ship to safety ef®cacy of clinical pro-ducts. However it is only through theclose communication between clini-cians and transfusion services and thetransfusionist with the manufacturersthat one is able to promote the appro-priate production strategies for achiev-ing the ultimate goal, reducing mortalityin thrombocytopenia platelet transfu-sion associated complications.

350 Transfus. Sci. Vol. 18, No. 3