Abstract

Light transmission aggregometry (LTA), which was independently developed in 1962 by Born 1 and O'Brien 2, is considered the gold standard for testing platelet function, because it provides important information that is essential for the diagnostic work-up of patients with platelet function defects. LTA measures the transmission of light through a sample of platelets in suspension (platelet-rich plasma [PRP], washed platelets or gel-filtered platelets), which increases when platelets are aggregated by an agonist. LTA is a time-consuming and technically challenging technique that is affected by many pre-analytical and analytical variables, and these must be carefully controlled for by expert personnel. For this reason, LTA should be performed only in specialized laboratories. Alternative methods to measure platelet aggregation in PRP or whole blood have been developed (e.g. impedance aggregometry, 96-well plate aggregometry, single platelet counting and flow cytometry) 3-7, several of which enable faster and more user-friendly study of the platelet response to aggregating agents. Despite these potential advantages, the majority of these techniques have not been widely adopted and, at variance with LTA, fail to provide important additional diagnostic information that can identify a defect of platelet function, such as platelet shape change, the occurrence of secondary wave of aggregation or platelet deaggregation. Although popular and widely used, the LTA technique is not standardized 8, despite the fact that guidelines have been published 9-11. Surveys organized by proficiency testing organizations identified variations in LTA practices and the need for guidelines to standardize LTA 12-14. In July 2006, the Platelet Physiology Subcommittee of the ISTH formed a Working Party of 11 experts with the aim of producing a series of consensus recommendations for standardizing LTA. As a first step in LTA guideline development, the Working Party organized the largest and most detailed worldwide survey on LTA methodology 15. The survey confirmed the very high variability in LTA practices worldwide, indicating that methodological standardization is necessary. The information gathered in the survey contributed to the development of ISTH methodological guidelines for LTA, which are the subject of the present report by the Working Party. Clinical or laboratory guidelines are traditionally based upon reviewing the evidence of the medical and scientific literature. However, it is unfortunately common that the literature does not provide definitive answers for a variety of reasons, including absence of evidence, and low quality and/or contradictory evidence. It was clear to the experts of the Working Party that not enough relevant studies had been published that compared different methodologies used to perform LTA studies, which would have helped to develop evidence-based guidelines. The Working Party therefore used a formal consensus method (the RAND methodology) 16 to develop its recommendations. The RAND method – developed by the RAND Corporation in the 1980s – is intended to obtain a formal consensus among expert groups about the appropriateness of healthcare interventions, particularly when scientific evidence is absent, scarce and/or heterogeneous 9. A series of statements about LTA practice were formulated by the chairman (MC) and the co-chairman (ADM) of the Working Party. For each statement, a form was prepared, which each member of the Working Party scored for appropriateness from 1 (completely inappropriate) to 9 (fully appropriate). Ballots were blinded to the other members. The extreme scores (highest and lowest) were discarded, the median of the remaining scores was calculated and the area containing the majority of the ballots defined the classification of each statement about LTA practice: inappropriate (scores 1–3), uncertain (4–6) and appropriate (7–9). After the first RAND run, statements were changed according to the comments/suggestions that the reviewers wrote on their RAND forms. A second RAND run was organized to score the revised statements. Recommendations were then presented at the ISTH SSC Meeting in Vienna (2008) and discussed with the audience. Based on the comments and suggestions that were raised during that meeting, statements were revised and a third RAND run was organized among the experts of the Working Party. The revised recommendations were again presented at an ISTH SSC Meeting (Boston, 2009) and discussed with the audience. Based on the comments and suggestions that were raised during that meeting, statements were again revised and a fourth RAND run was organized among the experts of the Working Party. At this point, before final approval of the recommendations, it was decided that a formal, full literature review should be performed, in order to verify, where possible, the recommendations of the Working Party. The Medline electronic database was searched for relevant published studies between 1966 and November 2009, without any language restriction, using the following medical subject headings (MeSH) terms and free text terms: (i) ‘platelet aggregation’ OR ‘platelet agglutination’; and (ii) ‘lumiaggregometry’ OR ‘light transmission aggregometry’ OR ‘light transmittance aggregometry’ OR ‘platelet reactivity tests’ OR ‘platelet function tests’ OR ‘platelet function test’ OR ‘Born aggregation’. The search was supplemented by manually reviewing the reference lists from primary or review articles and by direct consultation among members of working group to identify additional relevant studies. All references were reviewed manually and data extracted using the following inclusion criteria: (i) studies that directly compared different methodological approaches in studying LTA, including case control, prospective cohort or retrospective cohort studies; and (ii) studies in which one of the outcomes of interest (i.e. the methodological approach to study LTA) was a prespecified endpoint. By this method 1830 potentially relevant studies were identified and screened for retrieval. The titles and abstracts of these were divided into five groups, and each group was reviewed by two panel members. Based on their assessments, 1803 studies were excluded, leaving 27 for detailed evaluation. After review of the full manuscripts by the panel members, fifteen further studies were excluded. These 12 manuscripts, coupled with two extra studies identified by manual review and by consultation among members of the working group, resulted in a total of only 14 relevant studies 17-30 that were used to help refine the RAND survey and write the present recommendations (Fig. 1). Disagreement among experts was resolved by consensus and, when necessary, by asking the opinion of the chairman (MC). Based on the information that was gathered from the selected 14 publications, six consensus statements were slightly changed and a final, fifth RAND run among the Working Party experts was organized. The final statements were presented and discussed at the ISTH Subcommittee Meeting in Cairo (2010) and it was agreed that these could form the basis of a consensus document, which is the subject of the present report by the Working Party. The recommendations of the Working Party on standardization of LTA include 70 statements, which have been grouped into eight sections: (i) clinical usefulness of LTA; (ii) pre-analytical variables; (iii) blood collection; (iv) preparation of PRP and platelet-poor plasma (PPP); (v) assessment of PRP quality; (vi) methodology; (vii) choice of agonists; and (viii) evaluation and reporting of results. The experts agreed that this is an area that still needs further studies and standardization. New, faster and more efficient point-of-care tests are available for this purpose, but LTA could be used when the new tests are not available. However, based on the negative results of three large randomized controlled trials 31-33, which used the point-of-care test VerifyNow P2Y12, and a large non randomized, non controlled study, which used LTA to tailor antiplatelet treatment with clopidogrel 34, convincing data for recommending monitoring antiplatelet therapy by laboratory tests, including LTA, are still lacking. Based on the above three recommendations, the following recommendations apply only to the use of LTA for diagnosing patients with bleeding disorders, in whom the presence of a platelet function disorder is suspected. The effects of light meals on the results of LTA studies are probably negligible. However, patients should not be studied after having meals associated with high fat content, to avoid the formation of chylomicrons in plasma, which will interfere with light transmission. It is impossible for some patients to stop all medications before being sampled for LTA studies. Blood samples for LTA studies should be collected using the following precautions to minimize platelet activation during the procedure. It is also suggested that if a tourniquet needs to be used, it should immediately be released as soon as blood collection begins. Although five experts declared their preference for one of the two suggested concentrations of sodium citrate, the final recommendation was that either concentration of sodium citrate is acceptable, as long as its use is consistent in each center. This recommendation was recently validated by Femia et al. 35, who demonstrated that centrifugation of blood samples at 200 × g (or 250 × g) for 10 min appears to be the best condition for preparing PRP for LTA studies, both in terms of the degree of contamination of PRP by other blood cells and of platelet reactivity. Similar results in terms of platelet reactivity were reported by Merolla et al. 36. Recent studies demonstrated that platelet counts in PRP within the range that is observed in PRP samples from subjects with normal platelet count in whole blood do not affect the results of LTA studies 19-22. Therefore, the common practice of adjusting the platelet count in PRP with autologous PPP is not recommended, because it is unnecessary and may impair the platelet responsiveness to agonists 21. Uncertainty remains over what is the best practice to follow when the platelet count in PRP exceeds about 600 × 109 L−1. More recently, a study showed that both methods – native and adjusted platelet count – are valid to assess a bleeding disorder 37. Abnormalities of platelet aggregation were more frequent using adjusted platelet count both in controls and patients 37. Platelet agonists should be properly stored and checked for stability. The following platelet agonists, at the indicated concentrations, should be used for diagnostic LTA studies: Reference to Horm collagen (the most widely used collagen preparation for LTA studies) is given as an example and is not intended to bind laboratories to the use of this commercial preparation. M. Cattaneo: concept and design of the study, grading of recommendations, analysis of the literature search, analysis and interpretation of the data, writing the manuscript, and final approval of the version to be published. C. Cerletti, P. Harrison, C. P. M. Hayward, D. Kenny, D. Nugent, P. Nurden, A. K. Rao, A. H. Schmaier, S. P. Watson: grading of recommendations, analysis of the literature search, analysis and interpretation of the data, revising the manuscript, and final approval of the version to be published. F. Lussana, M. T. Pugliano: literature search, analysis of the literature search, analysis and interpretation of the data, revising the manuscript, and final approval of the version to be published. A. D. Michelson: concept and design of the study, grading of recommendations, analysis of the literature search, analysis and interpretation of the data, revising the manuscript, and final approval of the version to be published. The authors state that they have no conflict of interests.