Article Text

Download PDFPDF

Original research
Predictive value of platelet reactivity unit (PRU) value for thrombotic and hemorrhagic events during flow diversion procedures: a meta-analysis
  1. Ebunoluwa Ajadi1,
  2. Shaowli Kabir2,
  3. Aaron Cook3,
  4. Stephen Grupke8,
  5. Abdulnasser Alhajeri4,8,
  6. Justin F Fraser4,5,8,9
  1. 1 University of Kentucky College of Medicine, Lexington, Kentucky, USA
  2. 2 Department of Statistics, College of Arts and Sciences, University of Kentucky, Lexington, Kentucky, USA
  3. 3 Department of Pharmacy Practice and Science, University of Kentucky College of Pharmacy, Lexington, Kentucky, USA
  4. 4 Department of Neurology, University of Kentucky, Lexington, KY, USA
  5. 5 Department of Neurological Surgery, University of Kentucky, Lexington, Kentucky, USA
  6. 8 Department of Radiology, University of Kentucky, Lexington, KY, USA
  7. 9 Department of Neuroscience, University of Kentucky, Lexington, KY, USA
  1. Correspondence to Dr Justin F Fraser, Neurological Surgery, University of Kentucky, Lexington 40536, USA; Jfr235{at}uky.edu

Abstract

Background and purpose Platelet function testing prior to flow diversion procedures, although initially heavily debated, has seen a substantial increase in its adoption to assess the risk of operative and perioperative thrombotic and hemorrhagic events. This meta-analysis was conducted to assess platelet function testing, particularly the VerifyNow Platelet Reactivity Unit (PRU) assay, for a relationship between the reported assay PRU value and thrombotic and hemorrhagic events.

Materials and methods The currently available literature (2013–2018) was surveyed with PubMed and Google Scholar searches. Included studies were those for which there were at least 30 cases during the study period, for which VerifyNow platelet reactivity unit values were obtained prior to the procedures and for which intraoperative and perioperative adverse events were noted. PRU value cut-offs ranging from >200 to >240 comprised the hyporesponse group while values ranging from <60 to <70 comprised the hyper-response group. The data were subject to statistical analysis to assess the relationship between PRU values and thrombotic and hemorrhagic events. The collected data were subsequently statistically analyzed to assess for publication bias.

Results The searches yielded 27 studies, of which 12 met the inclusion criteria for the meta-analysis. The meta-analysis included data from 1464 reported Pipeline cases. The study included 273 men and 1177 women with a mean age across the analyzed procedures of 58 years (range 25–85). After loading with antiplatelet medications, preprocedural platelet hyper-responsiveness was associated with a greater incidence of hemorrhagic events with an increased absolute risk of 12%, but showed no relationship with thrombotic events. Preprocedural platelet hyporesponsiveness was associated with a greater incidence of thrombotic events with an absolute risk of 15%, but showed no relationship with hemorrhagic events.

Conclusions VerifyNow PRU values that correspond to platelet hyporesponse or hyper-response to dual antiplatelet therapy are associated with a higher risk of thrombotic and hemorrhagic events, respectively. Thus, the PRU value may offer some predictive value for these events.

  • aneurysm
  • blood flow
  • device
  • flow diverter
  • stent
View Full Text

Statistics from Altmetric.com

Introduction

Platelet testing, particularly the use of VerifyNow (Accumetrics, San Diego, California, USA), prior to neurointerventional procedures has risen in recent years,1–3 although debate remains regarding the clinical utility of such testing.3–7 This increase in preprocedural testing has closely paralleled the rise in the use of flow diverters in the surgical management of anterior and posterior circulation aneurysms.8 With the primary concern for surgeons/interventionalists deploying these flow-diverted stents being thrombotic and hemorrhagic complications, preprocedural platelet testing has risen as a means to assess the risk of these complications prior to the placement of these flow diversion devices. Notably, the risk of these events, and subsequent rise in preprocedural testing, is born out of the pharmacogenetic variability of clopidogrel as a component of the standard dual antiplatelet therapy administered to patients prior to these procedures and consequent hemodynamic alterations.9

Significant variability in patient response to clopidogrel has been widely reported.1 2 This variability has been primarily attributed to interindividual polymorphisms in CYP2C19 as well as other cytochromes in the P450 family that are required to metabolize and convert clopidogrel into its active P2Y12 receptor inhibiting form. Other factors have been suggested to contribute to the variable pharmacokinetics of clopidogrel including variable plasma-protein binding, P2Y12 receptor polymorphisms, and non-specific environmental factors such as smoking.1–3 The consequence of this variability has been a difficulty in predicting and thus subsequently managing patients who require dual antiplatelet therapy for the prevention of stent thrombosis during interventional procedures. Platelet function tests, particularly point-of-care tests like VerifyNow, have been increasingly adopted as a means to assess patient responses, anticipate complications, and better manage these patients perioperatively.

This study aims to assess the currently available literature for a potential relationship between platelet function testing and complications and outcomes during flow diversion procedures. The study also aims to systematically evaluate the relationship between platelet reactivity unit (PRU) values obtained from the VerifyNow platelet function test and thrombotic and hemorrhagic events during flow diversion embolization procedures. Finally, the study aims to compare the testing-related procedural outcomes that are available in the literature across multiple centers, and in doing so provide insight as to specific demographic, procedural-related factors, and additional confounding variables that could positively or negatively influence the occurrence of negative procedural outcomes, particularly thrombotic and hemorrhagic events.

Methods

Search strategy

This study was conducted in accordance with the proposed standards for observational meta-analysis and in accordance with  PRISMA guidelines10 (see online supplementary file 1), with PubMed and Google Scholar employed as the primary search databases. The searches were executed to obtain results relating to the use of PRU during endovascular procedures, and then narrowed to the use of PRU values in flow diversion procedures. In order to attempt homogeneity, we limited the study to those reporting on the use of the Pipeline Flow Diverter (Medtronic, Minneapolis, Minnesota, USA). The key single search terms included those related to platelet testing including “platelet reactivity units”, ’PRU', ’P2y12', and ’VerifyNow'. Those relating to the Pipeline procedure included ’Pipeline', ’Pipeline embolization', ’Pipeline Embolization Device', and ’PED'. Combined key words/search terms yielding the most relevant results included ’pru[All Fields] AND Pipeline[All Fields]', ’p2y12[All Fields] AND Pipeline[All Fields]', ’verifynow[All Fields] AND Pipeline[All Fields]', and ’pipeline embolization'. Individual publications between the years 2013 and 2018 were included in this study, and no hand searching was required given the relative novelty of platelet testing for flow diversion procedures.

Inclusion/exclusion criteria

Publications included in the study were those with at least 30 cases under the period of study, for which VerifyNow PRU values were obtained prior to the flow diversion procedures and for which adverse events both intraoperatively and perioperatively were noted: in particular, hemorrhagic and thrombotic events related specifically to flow diversion with the Pipeline Embolization Device (PED). Priority was also given to studies in which patients were subdivided/categorized based on PRU values into hyporesponders and hyper-responders, with a PRU value >240 being designated as the threshold for hyporesponse and a PRU value <60 being designated as hyper-response. Studies were categorized based on institutionally reported thresholds for platelet response, with hyporesponders forming one group and hyper-responders comprising the other group. No exclusions were made on the basis of patient demographics, procedural times, or comorbidities.

Statistical analysis

Thrombotic and hemorrhagic outcome variables were subject to meta-analysis with overall rates reported and a 95% CI applied where appropriate. The random effects model was applied in the regression analysis of outcome variables included in this study given that the observed effect sizes varied from one study to the next. To assess for publication bias, a funnel plot was generated and the included studies were then subject to a regression analysis using the Egger’s test. Necessary adjustments were subsequently made using Duval and Tweedie’s Trim and Fill algorithm.

Results

Database and patient summary

The term ‘platelet reactivity unit’ yielded 347 records, while the term ‘pipeline embolization device’ yielded 580 records; 35 records were identified with simultaneous references to both terms. Of these 35 records, 27 publications with direct relevance to this study were identified between 2013 and 2018. Of the 27 publications, 12 publications were found to meet the inclusion criteria for the study. The most common reasons for exclusion were lack of reporting of preprocedural PRU values, small sample size, and studies in which Pipeline procedures were not performed as the primary neurovascular intervention. Pretreatment with the standard high-dose aspirin (~300 mg) and clopidogrel (75 mg) was reported in 91% of the included studies. All studies initiated pretreatment between 4 and 10 days prior to the procedure. Of those that reported, the final pretreatment prior to the Pipeline procedure was conducted within 24 hours in most cases. Preprocedural PRU values were collected within the 24 hours preceding the flow diversion procedure in all the cases in this study. Reported thrombotic and hemorrhagic events in this study consist of those which occurred both intraoperatively and perioperatively, with those that occurred perioperatively defined as occurring within 48 hours of the Pipeline procedure.

The current meta-analysis included data from 1464 reported Pipeline cases. This study included 273 men and 1177 women with a mean age of 58 years (range 25–85). The indications for the procedures were reported as anterior or posterior circulation aneurysms in all cases. Table 1 summarizes the currently available data that met the inclusion criteria for this study. For each of the included publications, intraoperative and perioperative hemorrhagic and thrombotic events have been correlated with their corresponding preprocedural PRU values, with patients primarily grouped as hyper-responders (ie, PRU <60) and hyporesponders (ie, PRU >240).11–14 We also included studies that self-defined their thresholds for platelet inhibition. For example, in some specific studies responders were defined as those with a PRU value of less than either 190, 200, or 208. For instance, Bender et al 15 applied a 200 PRU value cut-off whereas Tan et al 8 applied a 208 cut-off. For these studies, non-responders were defined by the authors as those with PRU values above these thresholds. Finally, an earlier study by Raychev et al 16 reports platelet response as percent inhibition, with <50% inhibition being designated as poor responders and values above 75% designated as over-responders. These designations were the standard of reporting results of the VerifyNow assay before the introduction of the PRU value. Studies were categorized based on institutionally reported thresholds for platelet response, with hyporesponders and poor responders forming one group (PRU >200–>240) and hyper-responders and over-responders (PRU <60–<70) comprising the other group.

Table 1

Summary of demographic, pretreatment platelet function results, and thrombotic and hemorrhagic complications during flow diversion procedures for studies that met the inclusion criteria for the current study

It is important to note that in the included studies which were published by the same author, in particular the studies published by Almandoz et al,11–13 the analyzed patient groups were not identical; however, there was some overlap between the study periods. Thus, as with all meta-analyses and similarly designed studies, there exists the potential that these groups, although not identical, shared some patients. To account for this, only the most recent publication by this group was included in the statistical analysis of patient outcomes.

Platelet inhibition response and complications

Overall, 60 thrombotic and 31 hemorrhagic events were reported across all studies of patients undergoing Pipeline procedures. Figure 1 shows the relationship between hyper-response and hemorrhagic complications. A preprocedural platelet hyper-response was associated with a greater incidence of hemorrhagic events with an increased absolute risk of 12% and showed no relationship to thrombotic events. Figure 2 shows the relationship between hyporesponse and thrombotic complications. A preprocedural platelet hyporesponse was associated with a greater incidence of thrombotic events with an absolute risk of 15% and showed no relationship to hemorrhagic events.

Figure 1

Summary of meta-analysis highlighting the proportion of hemorrhagic complications among patients who showed ahyper-response to VerifyNow antiplatelet testing. The Forest plot illustrates the relative weight assigned to each study for the calculation of this variable. Only studies with non-zero variance were included in this analysis.

Figure 2

Summary of meta-analysis highlighting the proportion of thrombotic complications among patients who showed a hyporesponse to antiplatelet testing. The Forest plot illustrates the relative weight assigned to each study for the calculation of this variable. Only studies with non-zero variance were included in this analysis.

Egger’s test

For both the hyper-response and hyporesponse groups, the weighted regression with multiplicative dispersion model was applied with SE employed as the predictor. For the hyper-response group, the test for funnel plot asymmetry (figure 3) yielded a t value of 4.6246 with 3 df and a p value of 0.02. For the hyporesponse group, the test for funnel plot asymmetry (figure 4) yielded a t value of 3.2034 with 8 df and a p value of 0.01. These results indicate that full symmetry was not observed, and thus there exists a potential publication bias.

Figure 3

Funnel plot to assess for publication bias among studies that reported a proportion of hemorrhagic complications among hyper-responders to platelet inhibition. Egger’s test was performed to check symmetry of the funnel plot; proportions and variances were used to conduct this test.

Figure 4

Funnel plot to assess for publication bias among studies that reported a proportion of thrombotic complications among hyporesponders to platelet inhibition. Egger’s test was performed to check symmetry of the funnel plot; proportions and variances were used to conduct this test.

Discussion

This study was conducted to review and analyze the current evidence regarding the use of the VerifyNow PRU assay for platelet testing prior to Pipeline procedures. In particular, this review and meta-analysis aimed to determine the relative validity of the use of preprocedural PRU values in predicting thrombotic and hemorrhagic events associated with the Pipeline procedure. The current evidence validates the utility of preprocedural values as a predictor of hemorrhagic and thrombotic events either during or after Pipeline procedures. All but one15 of the included studies showed a statistically significant relationship between high PRU values and thrombotic events and, conversely, lower PRU values and hemorrhagic events. A greater degree of statistical significance was observed in the studies that delineated hyper-responders or hyporesponders or, alternatively, responders and non-responders at more extreme PRU values. It is important to note, however, that this relative significance was not controlled for by sample population size. Of the studies included in this analysis, a 1:1.57 ratio was observed between hyporesponders and hyper-responders. On analysis, the hyporesponse group showed a greater propensity for its respective complications. The degree to which the applied inclusion criteria influenced these findings is certainly open to question, given that all confounding variables could not be controlled for due to the nature of this study—particularly procedural complexity, as indicated by procedural time, and anatomical location of the aneurysm. However, consistency in findings between the included studies, as well as the consistent demographic characteristics of the patients in this study, are suggestive of the generalizability of the results. The ratio of men to women, age of patients, and comorbidities were relatively consistent across the included studies. The pretreatment regimen was also consistent across studies with all studies pretreating with the standard 75 mg clopidogrel regimen and all but one study co-administering high-dose aspirin (300–325 mg). The atypical study14 used 81 mg aspirin prior to the procedure.

As previously highlighted, a survey of the currently available literature suggests a progressive emergence of platelet testing with a parallel increase in the use of the VerifyNow PRU assay. Indeed, in conducting this meta-analysis, more studies than not had conducted a platelet function test prior to the flow diversion procedure. Thus, exclusion on the basis of platelet testing was based primarily on the type of platelet function test employed—that is, VerifyNow versus other forms of platelet testing. The emergence of the VerifyNow assay and subsequent increase in its application, although initially attributable to convenience and ease of administration, has seen a shift towards increased use based on potential clinical utility in predicting thrombotic and hemorrhagic complications. A number of publications11–14 have pointed to other factors including a greater degree of standardization offered as a result of reporting PRU values and improvements in test-retest reliability.

This meta-analysis demonstrates the predictive value of the PRU value for thrombotic and hemorrhagic events and represents the first empirical demonstration of the extent of the relationship between the PRU value and thrombotic and hemorrhagic events during flow diversion procedures. In addition, this meta-analysis demonstrates the degree to which this effect exists and paves the way for future studies on other independent predictors of these complications.

Limitations of study

As with most studies of this kind, the use of multicenter, multi-study data implies a lack of complete control for confounding variables. This review thus limits it scope to a general survey of the platelet testing landscape, with the highlighted results serving to direct further studies and thus better inform clinical decision-making. As previously highlighted, the results from this study suggest a potential publication bias—that is, the potential that individual authors tended to report work that fit into the current paradigm or existing consensus. Although this potential bias represents a key limitation to this study, it is important to note that several other study characteristics limit this potential including the relative novelty of the VerifyNow PRU assay in the setting of flow diversion procedures, a lack of initial consensus on the utility of this test, and strict inclusion/exclusion criteria that do not factor any outcome variables. The slight variability in applied cut-offs also presents a key limitation to the interpretation of the results of this study. Thus, adoption of a universal standard for hypo- and hyper-response cut-offs is an important consideration in the future of flow diversion. Despite this limitation, our results highlight the general utility of establishing these cut-offs. Finally, aneurysm location and size data were not available in most of the included studies; as such, the potential impact of both of these variables on the results of this study remain uncertain.

Conclusion

This study highlights the predictive value of the PRU value for thrombotic and hemorrhagic events during flow diversion procedures and demonstrates the utility of categorizing patients into hyper-responders and hyporesponders to dual antiplatelet therapy based on their preprocedural values. Preprocedural PRU values that corresponded to either hyper-response or hyporesponse were associated with a statistically significant risk of hemorrhagic and thrombotic events, respectively. This study paves the way for further investigation into other independent predictors of hemorrhagic and thrombotic events during flow diversion procedures, as well as further investigation into the utility of the PRU value in other procedural contexts, perhaps with antiplatelet agents such as ticagrelor for which there is significantly less pharmacogenetic variability.

References

View Abstract

Footnotes

  • Contributors EA: collected, cleaned and analysed the data, designed the data collection tools and drafted and revised the paper. SK: designed the key statistical analysis tools and analyzed the data. AC, SG, AE: analysed the data and revised the draft paper. JFF: monitored and oversaw data collection and statistical analysis, drafted and revised the paper.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests JFF is an equity interest holder for Fawkes Biotechnology and a consultant for Stream Biomedical and Medtronic Neurovascular.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.