Edited by: Marina Panova-Noeva, Johannes Gutenberg University Mainz, Germany
Reviewed by: Avi Leader, Rabin Medical Center, Israel; Salvatore De Rosa, University of Catanzaro, Italy
This article was submitted to Atherosclerosis and Vascular Medicine, a section of the journal Frontiers in Cardiovascular Medicine
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Patients with coronary artery disease (CAD) face considerable risk of cardiovascular events (CVEs) despite the wide application of antiplatelet therapy (
The platelet population is not homogeneous, and RPs comprise the youngest population in the circulating platelet pool. Unlike mature platelets, nucleic acid-rich RPs have larger sizes, contain more dense granules, and have increased thrombotic activity (
However, unlike MPV, the conventional parameter of platelet turnover, which has been extensively studied (
We performed a systematic search on Web of Science, PubMed, Scopus, and Embase from inception to January 2020 without language restriction. The major search terms were as follows: reticulated platelets, immature platelet fraction, immature platelet count, platelet turnover, mortality, death, cardiovascular death, coronary artery disease, acute coronary syndrome, myocardial infarction, percutaneous coronary intervention, stroke, and revascularization. Specific search strategy is shown in the
Study types: observational studies (prospective or retrospective studies);
Population: adult patients were diagnosed with CAD;
Index prognostic factor: RP was the single biomarker that we reviewed for its prognostic value, and the laboratory parameters included either IPF% or IPC;
Comparator prognostic factors: the focuses were on the adjusted and unadjusted prognostic value of RPs. No comparator factor was considered when summarizing the unadjusted prognostic effect of RPs, since the unadjusted value of RPs was directly calculated from the absolute events. However, the adjusted value of RPs means the prognostic effect of RPs after adjusting for other conventional prognostic factors (
Outcomes: the clinical outcome of interest included cardiovascular death, non-fatal myocardial infarction (MI), non-fatal cerebrovascular accidents (CVA) (
Timing: the value was measured after CAD diagnosis.
Data search and study selection were performed by two investigators separately (Zhao Y and Lai R). Inconsistencies regarding the decision of the two reviewers were resolved by consensus. Any remaining disagreement was adjudicated to the senior consultant (Zhang Y). Current systematic review and meta-analysis were performed according to the Preferred Reporting Items for Systemic Reviews and Meta-analyses (PRISMA) and Meta-analyses of Observational Studies in Epidemiology (MOOSE) checklist (
Data from eligible studies were extracted by two investigators independently (Zhao Y and Lai R). The discrepancy between the two reviewers during data extraction was resolved by a third supervisor (Zhang Y). We identified the name of first author, year of publication, number of participants, type of cases, proportion of female, proportion of smoker, proportion of patients with hypertension, diabetes mellitus (DM), and dyslipidemia at baseline, antiplatelet medications at baseline, and follow-up duration.
The outcomes were cardiovascular death, myocardial infarction, cerebrovascular accidents, and unplanned revascularization. Mean or median RPs with associated interquartile in patients with and without outcomes were extracted. To assess the prognostic value of RPs after adjustment of conventional factors, we extracted the type of statistical model, adjusted risk estimates and corresponding 95% confidence interval (95% CI), and number of covariates from the original studies. To calculate the unadjusted risk estimates, the absolute number of CVEs was also extracted. The Engauge Digitizer was used to extract the absolute number of events from the published Kaplan–Meier analysis, if sufficient data for meta-analysis were not reported in the primary studies.
Based on Quality in Prognostic Factor Studies (QUIPS) checklists (
For studies with dichotomous outcomes, the level of RPs was classified as high or low according to cut-point reported by original studies. If studies dichotomized the level RPs into two categories, we used the original categories. If there were more than two categories, we recategorized into high and low groups for ease of pooling. Adjusted risk estimates of CVEs reported directly in the included studies were extracted and transformed to the natural logarithms, as described previously (
Considering the potential heterogeneity of prognostic studies as suggested previously (
Statistical analysis was performed using Review Manager 5.3 (The Cochrane Collaboration, Oxford, UK) and R 3.6.3 (R Core Team, Vienna, Austria). Two-sided tests with
We identified 1,159 publications from Scopus, 480 publications from Web of Science, 327 publications from PubMed, and 234 publications from Embase. After applying eligibility criteria to 1,569 studies, we included 6 observational studies into the systematic review and meta-analysis. A detailed flowchart of studies inclusion was provided in
Flow diagram for study selection. CAD, coronary artery disease; RPs, reticulated platelets; IPF%, immature platelet fraction; IPC, immature platelet count.
Included studies involved a total of 1,636 patients with CAD with 294 outcome events during follow-up. Four of the included studies took place in Europe (
Characteristics of included studies.
Country | Austria | Austria | Italy | Spain | Israel | USA |
Type of cases | Post-PCI | Post-PCI | Post-PCI | ACS | SCAD with DM | CAD |
Participants ( |
486 | 477 | 229 | 251 | 104 | 89 |
Laboratory parameter | IPF% | IPC | IPF% | IPF% | IPF% | IPC |
Methodologies of RPs | Sysmex XE-2100 | Sysmex XE-2100 | Sysmex XE-2100 | Sysmex XE-2100 | Sysmex XE-2100 | Sysmex XE-2100 |
Measuring time | 6–24 h after PCI | 6–24 h after PCI | 24–48 h after PCI | In the morning of the first day of hospitalization | N/A | Within 72 h of admission to the hospital |
Median of follow-up | 190 days | 5.8 years | 1 years | In-hospital admission | 2 years | 31 months |
Total CVEs ( |
86 | 110 | 22 | 31 | 15 | 30 |
Reported outcomes of interest | Death, MI, RVA, CVA | Death, MI, CVA | Death | Death | MI, RVA, CVA | Death, MI, RVA |
RPs in CVEs | 4% [2.9–5.4] | N/A | 3.7% [2.4–5.0] | 6.60% [4.20–10.80] | 4.57% | 5.3% [4.3–6.4] |
RPs in non-CVEs | 3.3% [2.4–4.7] | N/A | 2.8% [1.9–4.1] | 4.80% [3.10–6.95] | 2.53% | 3.7% [3.0–5.1] |
0.013 | N/A | 0.05 | 0.002 | <0.001 | 0.007 |
Five out of six studies reported that patients with CVEs had a significantly higher RPs level than patients without CVEs. As shown in
Age of patients ranges from 64 to 76 years. The proportion of female, smoker, and patients with hypertension, diabetes mellitus, and dyslipidemia are 30.6, 36.6, 75.3, 34.3, and 68.8%, respectively. Four studies only enrolled patients taking dual antiplatelet medications (
Baseline characteristics of study participants.
Tscharre et al. ( |
477 (100.0) | 64.3 | 149 (31.2) | 293 (61.4) | 410 (86) | 139 (29.1) | 376 (78.8) | Dual: aspirin and clodipogrel (100) |
Perl et al. ( |
104 (100.0) | 71.2 | 24 (23.1) | 13 (12.6) | 87 (83.7) | 104 (100.0) | 94 (90.4) | Single: aspirin or clodipogrel (100) |
López-Jiménez et al. ( |
251 (100.0) | 68 | 66 (26.3) | 63 (25) | 104 (41) | 82 (33) | 110 (44) | Aspirin (97) or clodipogrel (75) |
Cesari et al. ( |
229 (100.0) | 76 | 75 (32.8) | 69 (30.1) | 131 (57.2) | 57 (24.9) | 88 (38.4) | Dual: aspirin and clodipogrel (100) |
Ibrahim et al. ( |
89 (100.0) | 68.1 | 32 (40.0) | 22 (24.7) | 84 (94.4) | 37 (41.6) | 76 (85.4) | Dual: aspirin and clodipogrel (100) |
Freynhofer et al. ( |
486 (100.0) | 64 | 154 (31.7) | 140 (28.8) | 416 (85.6) | 140 (28.8) | 381 (78.4) | Dual: aspirin and clodipogrel (100) |
Two studies indicated moderate quality (
We found that asymmetry existed in the funnel plot. The trim and fill method and Egger's test identified publication bias (
Statistical models of included studies.
Tscharre et al. ( |
U&M Cox ph | IPC > 7,600/μL | HR, 1.693 (95% CI, 1.156, 2.481) | HR, 1.716 (95% CI, 1.152, 2.559) | 31 | MEA, VASP-P, MPV | Age, hyperlipidemia, peripheral artery disease, ACEI or ARB, DES |
Ibrahim et al. ( |
M Cox ph | IPC > 7,632/μL | HR, 4.65 (95% CI, 1.78, 12.16) | N/A | 7 | MPV, LTA | N/A |
Perl et al. ( |
M regression | IPF% > median | OR, 1.968 (95% CI, 1.1128–3.432) | N/A | 15 | MPV | Age, prior MI, anemia |
López-Jiménez et al. ( |
M regression | IPF% > 6.2% | OR, 2.42 (95% CI, 1.08, 5.43) | N/A | 5 | N/A | Admittance Killip |
Cesari et al. ( |
U&M regression | IPF% > 3.3% | OR, 2.83 (95% CI, 1.14–7.06) | OR, 4.15 (95% CI, 1.24–13.91) | 8 | MPV, LTA, H-IPF | H-IPF |
Freynhofer et al. ( |
U&M regression | IPF% > 3.35% | OR, 1.136 (95% CI, 1.001–1.288) | OR, 1.173 (95% CI, 1.040–1.324) | 41 | MEA, VASP-P, MPV | Troponin I, CRP, prior MI |
The absolute number of composite events (including fatal and non-fatal) with high and low RPs could be extracted from all included studies (i.e., from a two-by-two table format). For unadjusted risk of CVE, the pooled result of RR for composite CVEs was 2.26 (95% CI, 1.72–2.98) (
Forest plot for relative risk of composite CVEs in patients with high or low RP level.
Forest plot for relative risk of cardiovascular death in patients with high or low RP level.
Although four studies reported the non-fatal events as the outcomes, the absolute numbers were only available in three studies with the two-by-two table format. The separate meta-analysis suggested that RPs might not act as a prognostic predictor for myocardial infarction (RR, 1.93; 95% CI, 0.81–4.61) or cerebrovascular events (RR, 2.08; 95% CI, 0.91–4.75) (
Forest plot for relative risk of non-fatal CVEs in patients with high or low RP level.
Two studies used the Cox proportional hazard analysis and reported adjusted hazard ratios (HR) (
Five out of six studies (
After transforming the adjusted risk estimates, the summarized result of adjusted risk estimates was 2.00 (95% CI, 1.30, 3.08) (
Forest plot for adjusted risk estimates of composite CVEs in patients with high or low RP level.
We performed meta-regression and subgroup analysis to explore the source of heterogeneity of adjusted risk estimates across studies. As shown in
Meta-regression.
Like the results of meta-regression, the test for subgroup differences in “number of adjusted covariates” subgroup revealed there is a statistically significant subgroup effect (
Subgroup analyses on composite cardiovascular events.
Laboratory parameters of RPs | IPF% | 4 | 1,070 | 0%, |
72%, |
1.84 (1.07–3.16) |
IPC | 2 | 566 | 72%, |
2.56 (0.98–6.66) | ||
Follow-up | ≥2 years | 3 | 670 | 0%, |
44%, |
2.14 (1.36–3.35) |
<2 years | 3 | 966 | 73%, |
1.92 (0.88–4.17) | ||
No. of adjusted covariates | ≥10 | 3 | 1,067 | 83.3%, |
71%, |
1.47 (1.01–2.12) |
<10 | 3 | 569 | 0%, |
3.35 (1.93–5.81) |
We examined the association between circulating RP levels and cardiovascular outcomes in patients with CAD. The major findings were as follows: (1) The circulating levels of RPs were significantly higher in patients with CVEs, as reported by original studies. (2) In the analysis of adjusted risk estimates, the summarized results demonstrated that increased levels of RPs were associated with a higher risk of CVEs. Moreover, included studies also measured other platelet function tests (including LTA, MEA, and MPV), but none could independently predict CVEs, like RPs. Meta-regression and subgroup analysis demonstrated the different number of adjustment factors in the original studies was the source of heterogeneity, and the prognostic effect of RPs was more significant within groups having follow-up longer than 2 years and expressing laboratory results as IPF%. Publication bias were identified using funnel plot, Egger's test, and trim-and-fill method. (3) The meta-analysis of unadjusted risk estimates also confirmed the prognostic value of RPs for predicting the composite CVEs, cardiovascular death, and RVA. However, RPs might not be a good predictor for MI or CVA.
Besides other conventional risk factors like age and sex, platelets also play a pivotal role in cardiovascular diseases. High on-treatment reactivity might arise due to platelet turnover and lead to poor prognosis of patients with CAD. Our work found that RPs, the novel parameter of platelet turnover, might have the potential as an independent predictor for CVEs. A broad array of platelet turnover tests, namely MPV is already available in clinical practice. Previous meta-analysis and subgroup analysis reported MPV could act as a prognostic marker in patients with ACS other than patients that underwent post-PCI or SCAD (
The addition of RPs into current prediction models may improve the risk stratification for adverse cardiovascular events in patients with CAD. Currently, there are several prediction models available to enhance prognostication in the management of patients with CAD. For example, the Global Registry of Acute Coronary Event (GRACE) is recommended to predict mortality for patients with ACS by applying risk factors, cardiac biomarkers, and electrocardiograms. As summarized extensively, GRACE score's prognostic value is not perfect, and incremental effects after adding other laboratory parameters have been reported (
Monitoring of RP levels may also help assess the response to antiplatelet therapies. Included studies reported relatively uniform antiplatelet medications, aspirin and clopidogrel. Aspirin constitutes the critical component of secondary prevention of CAD and is complemented by clopidogrel in patients undergoing PCI (
Given the advent of a fully automated cell analyzer, the values of RPs become simple to obtain and easy to interpret from automatic cell counters (
Overall, included studies indicated moderate or high quality, and all of them were performed in developed countries. Study objectives and populations were clearly specified, and the valid and reliable measurement of RPs was also reported. All studies determined RPs before cardiovascular outcomes. All studies reported adjusted risk estimates based on appropriate statistical models. However, some limitations must be considered. First, the heterogeneity test of adjusted prognostic values suggested substantial heterogeneity across studies, which might be explained by the different adjustment of prognostic covariates. In addition, the sample size in primary studies are relatively small since RP measurement by the automated analyzer has only been employed for a short time in clinical practice. The identified publication bias also limited the robustness of the results. Therefore, the results should be interpreted with great caution. Based on standardized measurement of RPs, especially IPF%, large-scale observational studies with long follow-up duration and reliable confounding adjustments are needed for drawing a firm conclusion on the prognostic value of RPs in predicting adverse CVEs.
Our work suggested circulating level of RPs of patients with CAD could be a prognostic factor for CVEs in the future, even after adjustment of conventional covariates. Further large-scale studies with long follow-up duration are still necessary for drawing a firm conclusion of the prognostic value of RPs in patients with CAD.
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
YZhang and DS: conceptualization and supervision. YZhao and RL: data curation, formal analysis, investigation, and methodology. YZhao and YZhang: project administration. YZhao: software and writing (original draft). RL, YZhang, and DS: writing (review and editing). All authors contributed to the article and approved the submitted version.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank Dr. Jerry a professional English editor for his language help and writing assistance. We thank Hua Qu from Xiyuan Hospital, China Academy of Chinese Medical Sciences for his advice/help on the methodology and statistics used.
The Supplementary Material for this article can be found online at: