Introduction
Platelets are anucleate blood cells which are primarily produced in the bone marrow
in a process called thrombopoiesis. They derive from megakaryocytes through thrombopoietin–thrombopoietin
receptor interactions which induce the formation of pro-platelets.[1] Platelets are implicated in hemostasis and arterial thrombosis and hence significantly
contribute to development and exacerbation of cardiovascular disease. In that light,
acute myocardial infarction (MI) is characterized by erosion or rupture of atherosclerotic
plaques inside coronary arteries leading to thrombus formation (atherothrombosis)
which may result in obstruction of blood flow and subsequent ischemia of downstream
located tissue.[2]
[3]
[4] As a consequence, targeting platelet activity is a cornerstone in acute coronary
syndrome (ACS)/MI treatment.[5] Apart from promoting only thrombus formation, platelets can trigger acute coronary
events also by interacting with leukocytes, endothelial cells, and the coagulation
system resulting in thromboinflammation.[1]
[6]
While a higher platelet count on presentation was found to be associated with adverse
clinical outcomes,[7]
[8]
[9]
[10]
[11] studies that specifically examined the impact of platelet counts on myocardial ischemia
and recovery together with outcome in patients with acute ST-segment elevation myocardial
infarction (STEMI) are scarce. Therefore, we here explore the association of platelet
counts with infarct size using serial single-photon emission computed tomography (SPECT)
imaging and long-term mortality in patients with STEMI undergoing primary percutaneous
coronary intervention (PPCI).
Methods
Study Design
Details of the study patients were described earlier.[12]
[13]
[14] By design, the study represents a retrospective analysis. In brief, between January
2002 and December 2007, patients with STEMI undergoing PPCI and serial scintigraphic
imaging at two tertiary cardiac care centers (Deutsches Herzzentrum Muünchen and Klinikum
rechts der Isar, both Technical University of Munich, Munich, Germany) were included
in this study. The diagnosis of STEMI was based on chest pain lasting ≥20 minutes
and persistent ST-segment elevation ≥1 mm in at least two extremities or ≥2 mm in
at least two chest leads or new onset of left bundle branch block. As reported recently,[13] 200 out of 1,406 STEMI patients were excluded because the time-of-day at symptom
onset was not clearly documented. Hence, the remaining cohort of 1,206 patients was
included into this analysis with additional information (e.g., symptom onset) for
further analysis.[12] Two patients were excluded because admission platelet measurements were not available.
To exclude patients with severe thrombocytopenia, six patients with platelet counts
below 100 [109/L] on admission were excluded. Finally, 1,198 STEMI patients with serial scintigraphic
data were included in this study. All patients gave written informed consent for PPCI
and imaging procedures. The study protocol was approved by the institutional ethics
committee (454/21 S-KH) and conforms to the Declaration of Helsinki.
Angiography and PPCI
The culprit lesion in the infarct-related artery was identified during coronary angiography
by the presence of acute occlusion, intraluminal filling defects (or thrombus), ulcerated
plaques with contrast-filled pockets protruding into the plaque with or without delayed
contrast wash-out, extraluminal contrast, dissection, or intraluminal flaps. Coronary
artery disease in non-culprit lesions was defined as coronary stenosis of ≥50% lumen
obstruction.[13] Left ventricular ejection fraction (LV-EF) on admission (baseline) and after 6 months
was measured on left ventricular angiograms using the area-length method.[13] Unfractionated heparin was used for periprocedural anticoagulation.[13] The antithrombotic regime included clopidogrel, a loading dose of 600 mg, and aspirin
325 to 500 mg.[13] Chronic antithrombotic therapy consisted of clopidogrel, 150 mg until discharge
(no more than 3 days) mostly followed by 75 mg/day for ≥1 month and aspirin 200 mg/day
indefinitely.[13] Few patients were treated with ticlopidine (250 mg twice/day) instead of clopidogrel
and aspirin 200 mg/day indefinitely. If indicated—mostly due to new onset of atrial
fibrillation—patients were treated with phenprocoumon in combination with either aspirin
and clopidogrel or aspirin and ticlopidine, respectively.[12]
Measurement of Myocardial Area at Risk and Final Infarct Size Using SPECT
99mTc-sestamibi SPECT imaging studies were performed as described previously.[12]
[13]
[14] In brief, SPECT imaging was performed twice in each patient at predefined time points
using the following protocol: First measurement: 99mTc-sestamibi (27 mCi [1,000 MBq]) was injected intravenously before PPCI and
imaging was performed 6 to 8 hours afterward to assess the perfusion defect. This
estimates the myocardial area at risk. Second measurement: 99mTc-sestamibi was injected intravenously 7 to 14 days after PPCI and imaging was
performed 6 to 8 hours afterward. This estimates the final infarct size. Perfusion
defects were defined as <50% uptake of 99mTc-sestamibi and were expressed as percentage
of the left ventricle.[12]
[13]
[14] Myocardial salvage index (i.e., relative salvage) was calculated as initial myocardial
area at risk minus final infarct size divided by initial myocardial area at risk.[12]
[13] The myocardial salvage index represents the proportion of initial myocardial area
at risk salvaged by reperfusion therapy. All measurements were performed by investigators
who were not aware of the clinical or angiographic data.[12]
[13]
[14]
Laboratory Data, Medical History, and Definitions
Laboratory measurements were performed daily at the institute of laboratory medicine
of our hospital and extracted from patients' charts up to 10 days of hospitalization.
Baseline platelet count refers to the first available measurement of platelet count
recorded on admission. For the purpose of this analysis, patients were divided into
three groups according to the tertiles of platelet count on admission: a group with
platelet count in the 1st tertile (T1 group), a group with platelet count in the 2nd
tertile (T2 group), and a group with platelet count in the 3rd tertile (T3 group).
Normal platelet count was defined as 140 to 400 [109/L]. Creatine kinase myocardial band (CK-MB) was measured daily and peak levels were
defined as the highest value obtained during hospitalization. CK-MB was considered
an enzymatic estimate of infarct size. Renal function was assessed by calculating
the creatinine clearance according to the Cockroft–Gault formula.
Study Outcomes and Follow-up
The primary endpoint was 1-year all-cause mortality. Secondary endpoints were: myocardial
salvage index, LV-EF (at 6 months), non-fatal MI (at 1- and 5-year follow-up), 5-year
all-cause mortality, and MACE (1- and 5-year follow-up). As a standard practice in
our institutions at the time of patient's recruitment, a repeat coronary angiography
at 6 months after the index procedure was scheduled. The 6-month angiograms were used
for the assessment of the LV-EF at this time point. Nonfatal MI was diagnosed based
on the development of new abnormal Q waves in two or more contiguous chest or two
or more adjacent extremity leads, or an elevation of CK-MB more than two times (more
than three times for 48 hours after a percutaneous coronary intervention [PCI] procedure)
the upper limit of normal in the presence of ischemia symptoms.[13] Follow-up information was obtained by staff members who were not aware of the clinical
data via phone calls 30 days after PCI, 1 year after PCI, and yearly thereafter.[12]
[13] Data on mortality were obtained from hospital records, death certificates, or phone
contact with patients' relatives or referring physicians.[12]
[13]
Statistical Analysis
Continuous data are shown as mean ± standard deviation or median with 25th to 75th
percentiles and 10th to 90th percentiles depending on the distribution of normality
and compared with one-way ANOVA, Kruskal–Wallis test (plus Dunn's multiple comparisons
test).[12]
[13] Discrete variables were shown as proportions (percentages) and compared with chi-square
test.[12]
[13] Long-term clinical outcomes were assessed using the Kaplan–Meier method and log-rank
test. Multivariable Cox proportional hazards model was used to assess the association
between baseline platelet count and both all-cause mortality and MACE as described
previously.[12] Hazard ratios (HRs) were shown with 95% confidence intervals (CIs). Age, heart rate
(on admission), systolic blood pressure (on admission), creatinine clearance, sex,
type II diabetes, Killip class on admission, previous MI, number of diseased vessels,
platelet counts on admission were entered into the model. In the second model, leukocyte
count and CRP level on admission or on day 1 were added alongside other baseline variables.
A two-sided p-value of <0.05 was considered to indicate statistical significance. IBM SPSS Statistics
version 29 and GraphPad Prism 9 were used for statistical analysis and visualization
of data.
Results
Baseline Data
This study included 1,198 patients who were categorized in groups according to the
tertiles of platelet count: a group with platelet count in the 1st tertile (T1 group;
platelet count, 102–206 [109/L]; n = 402), a group with platelet count in the 2nd tertile (T2 group; platelet count,
207–259 [109/L]; n = 396), and a group with platelet count in the 3rd tertile (T3 group; platelet count,
260–921 [109/L]; n = 400). Median platelet counts on admission were 183 [109/L] in the T1 group, 232 [109/L] in the T2 group, and 294 [109/L] in the T3 group ([Fig. 2A]). Platelet count change over the hospital course is shown in [Supplementary Fig. S1A] (available in the online version only) for all patients and in [Supplementary Fig. S1B] (available in the online version only) for individual groups. Baseline characteristics
of patients in groups T1 to T3 are shown in [Table 1]. The baseline variables differed between the groups with respect to age, sex, current
smoking, family history for coronary artery disease (CAD), creatinine clearance on
admission, previous coronary artery bypass graft surgery (CABG), and location of the
culprit lesion. Data on medication on admission and at discharge are included in [Supplementary Table S1] (available in the online version only). Ninety-two patients in T1 (22.9%), 70 patients
in T2 (17.7%), and 58 patients in T3 (14.5%) were on aspirin; 4 patients in T1 (1.0%),
4 patients in T2 (1.0%), and 3 patients in T3 (0.8%) were on clopidogrel; and 4 patients
in T1 (1.0%), 3 patients in T2 (0.8%), and 2 patients in T3 (0.5%) were on both drugs
on admission. Most patients received aspirin and clopidogrel after PPCI, since other
potent P2Y12 inhibitors were not available at the time when the study was conducted.
The antiplatelet treatment did not differ on admission or at discharge between the
three groups. All patients were advised to take clopidogrel for 12 months and to continue
aspirin indefinitely after the index event.
Table 1
Baseline and procedural characteristics
|
T1 (102–206 [103/L])
|
T2 (207–259 [103/L])
|
T3 (260–921 [103/L])
|
p-Value
|
|
n = 402 (33.6%)
|
n = 396 (33.1%)
|
n = 400 (33.4%)
|
|
Platelets on admission, mean [103/L] (SD)
|
176.70 (22.94)
|
231.78 (14.71)
|
309.96 (59.16)
|
<0.001
|
|
Age, years, mean (SD)
|
64.5 (12.3)
|
62.2 (13.3)
|
60.1 (13.0)
|
<0.001
|
|
Male sex, n (%)
|
344 (85.6)
|
306 (77.3)
|
264 (66.0)
|
<0.001
|
|
Diabetes, n (%)
|
83 (20.6)
|
73 (18.4)
|
72 (18.0)
|
0.592
|
|
BMI, mean in kg/m2
(SD)
|
26.8 (3.6)
|
26.9 (3.8)
|
26.4 (4.3)
|
0.256
|
|
Hypercholesterolemia, n (%)
|
208 (51.7)
|
212 (53.5)
|
227 (56.8)
|
0.354
|
|
Hypertension, n (%)
|
286 (71.1)
|
270 (68.2)
|
285 (71.3)
|
0.562
|
|
Current smoking, n (%)
|
154 (38.3)
|
169 (42.7)
|
190 (47.5)
|
0.031
|
|
Family history, n (%)
|
139 (34.6)
|
171 (43.2)
|
166 (41.5)
|
0.031
|
|
CrCl, (mL/min)
mean (SD)
|
83.9 (30.7)
|
89.2 (34.9)
|
89.6 (36.1)
|
0.031
|
|
No. of affected vessels
|
0.204
|
|
1, n (%)
|
128 (31.8)
|
149 (37.6)
|
149 (37.3)
|
|
2, n (%)
|
124 (30.8)
|
119 (30.1)
|
130 (32.5)
|
|
3, n (%)
|
150 (37.3)
|
128 (32.3)
|
121 (30.3)
|
|
Previous MI, n (%)
|
61 (15.2)
|
47 (11.9)
|
41 (10.3)
|
0.098
|
|
Previous CABG, n (%)
|
17 (4.2)
|
15 (3.8)
|
5 (1.3)
|
0.032
|
|
Time to admission, hours, mean (SD)
|
6.9 (6.2)
|
6.3 (5.7)
|
6.7 (6.0)
|
0.303
|
|
Door to balloon, hours, mean (SD)
|
1.4 (0.8)
|
1.4 (0.9)
|
1.4 (0.8)
|
0.335
|
|
Killip class
|
0.464
|
|
I, n (%)
|
297 (73.9)
|
301 (76.0)
|
291 (72.8)
|
|
II, n (%)
|
80 (19.0)
|
73 (18.4)
|
79 (19.8)
|
|
III, n (%)
|
13 (3.2)
|
14 (3.5)
|
11 (2.8)
|
|
IV, n (%)
|
12 (3.0)
|
8 (2.0)
|
19 (4.8)
|
|
Baseline TIMI flow, n (%)
|
0.280
|
|
0, n (%)
|
188 (46.8)
|
182 (46.1)
|
194 (48.5)
|
|
1, n (%)
|
36 (9.0)
|
57 (14.4)
|
43 (10.8)
|
|
2, n (%)
|
104 (25.8)
|
87 (22.0)
|
86 (21.5)
|
|
3, n (%)
|
74 (18.4)
|
69 (17.5)
|
77 (19.3)
|
|
Type of PCI, n (%)
|
0.482
|
|
PTCA, n (%)
|
45 (11.2)
|
58 (14.6)
|
58 (14.5)
|
|
Stenting, n (%)
|
357 (88.8)
|
338 (85.4)
|
342 (85.5)
|
|
No reflow, n (%)
|
67 (16.7)
|
49 (12.4)
|
46 (11.5)
|
0.073
|
|
LV-EF, %, mean (SD)
|
48.8 (11.7)
|
50.0 (11.1)
|
48.4 (11.4)
|
0.121
|
|
Infarct vessel (culprit lesion)
|
0.011
|
|
Left main, n (%)
|
3 (0.7)
|
1 (0.3)
|
0 (0)
|
|
LAD, n (%)
|
163 (40.5)
|
171 (43.2)
|
201 (50.3)
|
|
LCX, n (%)
|
75 (18.7)
|
60 (15.2)
|
67 (16.8)
|
|
RCA, n (%)
|
150 (37.3)
|
156 (39.4)
|
131 (32.8)
|
|
CABG, n (%)
|
11 (2.7)
|
8 (2.0)
|
1 (0.3)
|
Abbreviations: BMI, body mass index; CABG, coronary artery bypass graft surgery; CrCl,
creatinine clearance; LAD, left anterior descending coronary artery; LCX, left circumflex
coronary artery; LV-EF, left ventricular ejection fraction; MI, myocardial infarction;
PCI, percutaneous coronary intervention; PTCA, percutaneous transluminal coronary
angioplasty; RCA, right coronary artery; SD, standard deviation; T1, tertile 1; T2,
tertile 2; T3, tertile 3; TIMI flow, thrombolysis in myocardial infarction flow.
Fig. 1 In patients with ST-elevation myocardial infarction, high-range blood platelet counts
on admission are associated with larger infarct size and reduced left ventricular
(LV)-function. Both low- and high-range blood platelet counts on admission are associated
with long-term mortality compared to patients with mid-range platelet counts.
Fig. 2 (A) Platelet counts on admission (tertile-based analysis). (B) Initial myocardial area at risk (% of the left ventricle) assessed by first scintigraphic
imaging prior to primary percutaneous coronary intervention (PPCI). (C) Final infarct size assessed by second scintigraphic imaging 7–14 days after PPCI
(% of the left ventricle). (D) Myocardial salvage index in different tertiles of platelet count on admission. Data
are median with 25th–75th percentiles (boxes), 10th–90th percentiles (bars), and values
outside the given percentiles (dots). LV, left ventricle; T1, tertile 1; T2, tertile
2; T3, tertile 3.
Baseline Platelet Counts and Infarct Size
Data on scintigraphic- and blood-based measures of infarct sizes are presented in
[Fig. 2B–D] and [Supplementary Fig. S2A, B] (available in the online version only). Myocardial area at risk or initial perfusion
defect before PPCI (median) was 22.0% (interquartile range [IQR]: 12.0–39.8%), 21.0%
(IQR: 11.0–37.1%), and 26.0% of the left ventricle (IQR: 14.9–45.0%) in T1 to T3 groups,
respectively. Patients in the T3 group showed the largest area at risk in comparison
to patients in the lower platelet count tertiles on admission ([Fig. 2B]). The infarct sizes in the 7- to 14-day scintigraphy (median) were 10.0% (IQR: 2.0–21.0%),
9.0% (IQR: 2.0–20.7%), and 12.0% of the left ventricle (IQR: 3.0–27.3%) in T1 to T3
groups, respectively. Again, the T3 group (T3) showed largest infarct sizes ([Fig. 2C]). In line with these findings, peak CK-MB and troponin T values, both enzymatic
estimates of infarct sizes, were highest in T3 ([Supplementary Fig. S2A, B] [available in the online version only]). Peak CK-MB levels (median) were 107.5 U/L
(IQR: 53.0–213.0 U/L) in the T1 group, 102.5 U/L (IQR: 41.0–193.0 U/L) in the T2 group,
and 119.0 U/L (IQR: 61.0–272.0 U/L) in the T3 group. Peak troponin T levels (median)
were 3.68 ng/mL (IQR: 1.32–7.04 ng/mL) in the T1 group, 3.0 ng/mL (IQR: 1.43–0.02 ng/mL)
in the T2 group, and 4.61 ng/mL (IQR: 1.88–9.29 ng/mL) in the T3 group. The myocardial
salvage index (median) was unchanged in between groups: 0.50 (IQR: 0.23–0.79), 0.53
(IQR: 0.27–0.81), and 0.48 (IQR: 0.26–0.80) in patients from groups T1, T2, and T3,
respectively ([Fig. 2D]).
Fig. 3 Left ventricular ejection fraction (LV-EF) at (A) baseline, (B) 6 months, and (C) comparison between baseline and 6 months according to platelet count tertiles. Data
are median with 25th–75th percentiles (boxes), 10th–90th percentiles (bars), and values
outside the given percentiles (dots). T1, tertile 1; T2, tertile 2; T3, tertile 3.
Baseline Platelet Counts and Left Ventricular Function
An angiographic assessment of the LV-EF during the index procedure was available in
1,140 (95.2%) patients: 378 (94.0%) patients in the T1 group, 381 (96.2%) patients
in the T2 group, and 381 (95.3%) patients in the T3 group. LV-EF (median) was 49.65%
(IQR: 42.0–56.0%), 50.51% (IQR: 44.0–58.0%), and 49.0% (IQR: 41.0–56.0%) in the T1,
T2, and T3 groups, respectively ([Fig. 3A]). LV-EF was lowest in the T3 group. LV-EF measurements at 6-month follow-up were
available in 468 (39.1%) of the patients: 144 (35.8%) patients in the T1 group, 152
(38.4%) patients in the T2 group, and 172 (43.0%) patients in the T3 group. At 6 months,
LV-EF values (median) were 60.75% (IQR: 52.0–67.85%), 61.40% (IQR: 50.30–68.50%),
and 56.0% (IQR: 47.0–67.0%) in the T1, T2, and T3 groups, respectively ([Fig. 3B]). LV-EF was lowest in patients of the T3 group. Data on LV-EF both at baseline and
6-month follow-up were available for 409 patients (34.1%): 102 (25.4%) patients in
the T1 group, 145 (36.6%) patients in the T2 group, and 162 (40.5%) patients in the
T3 group. A significant recovery (improvement compared with baseline values) of the
LV-EF at the 6-month follow-up angiography occurred in all groups regardless of the
platelet count on admission ([Fig. 3C]).
Baseline Platelet Counts and Clinical Outcomes
The median follow-up was 1,396 days. One- and 5-year follow-up for all-cause mortality
was available for 1,069 (89.2%) and 449 (37.5%) patients, respectively. The primary
endpoint (death of any cause at 1 year) occurred in 43 patients: 16 deaths in the
T1 group, 5 deaths in the T2 group, and 22 deaths in the T3 group (Kaplan–Meier estimates
of 1-year mortality: 4.2, 1.3, and 5.6%, respectively; log-rank test p = 0.006; [Fig. 4A]). At 5 years, deaths of any cause occurred in 103 patients: 39 deaths in the T1
group, 23 deaths in the T2 group, and 41 deaths in the T3 group (Kaplan–Meier estimates
of 5-year mortality: 13.1, 8.0, and 12.4%, respectively; log-rank test p = 0.057; [Fig. 4B]). Mortality was lowest in the T2 group at both 1- and 5-year follow-up.
Fig. 4 Kaplan–Meier curves of 1-year (A) and 5-year (B) all-cause mortality according to platelet count tertiles. T1, tertile 1; T2, tertile
2; T3, tertile 3.
At 1 year, MACE occurred in 300 patients: 98 events in the T1 group, 82 events in
the T2 group, and 120 events in the T3 group (Kaplan–Meier estimates of 1-year MACE:
25.7, 21.3, and 30.8%, respectively; log-rank test, p = 0.008; [Supplementary Fig. S3A] [available in the online version only]). At 5 years, MACE occurred in 371 patients:
130 events in the T1 group, 102 events in the T2 group, and 139 events in the T3 group
(Kaplan–Meier estimates of 5-year MACE: 37.9, 29.2, and 37.8%, respectively; log-rank
test, p = 0.017; [Supplementary Fig. S3B] [available in the online version only]). The incidence of MACE was lowest in the
T2 group.
At 1-year, non-fatal MIs occurred in 31 patients: 12 MIs in the T1 group, 6 MIs in
the T2 group, and 13 MIs in the T3 group (Kaplan–Meier estimates of 1-year mortality:
3.2, 1.5, and 3,3%, respectively; log-rank test, p = 0.166; [Supplementary Fig. S4A] [available in the online version only]). At 5 years, MIs occurred in 40 patients:
16 MIs in the T1 group, 8 MIs in the T2 group, and 16 MIs in the T3 group (Kaplan–Meier
estimates of 1-year mortality: 4.5, 2.4, and 4.7%, respectively; log-rank test, p = 0.199; [Supplementary Fig. S4B] [available in the online version only]). No difference was observed in between groups
regarding the occurrence of MIs.
The association of baseline platelet counts with 1- and 5-year all-cause mortality
and MACE was adjusted using Cox proportional hazard model (see methods for variables
entered into the model).
Patients in the T1 and T3 groups of platelet count had a higher adjusted risk for
all-cause death compared with patients of the T2 group (reference group): adjusted
HR, 3.404 (1.226–9.542), p = 0.019 and 3.549 (1.228–9.782), p = 0.014, respectively, at 1 year. At 5-year follow-up, mortality was numerically
higher in patients of the T1 and T3 groups compared with the T2 group (reference group):
adjusted HR, 1.628 (0.960–2.759), p = 0.071 and 1.708 (0.996–2.926), p = 0.052, respectively; [Tables 2] and [Supplementary Table S2] [available in the online version only]).
Table 2
The association of platelet count on admission with 1-year and 5-year all-cause mortality
after adjustment in the multivariable Cox proportional hazard model
|
1-year mortality
HR (95% CI)
|
p-Value
|
5-year mortality
HR (95% CI)
|
p-Value
|
|
Tertile 1
|
3.404 (1.226–9.542)
|
0.019
|
1.628 (0.960–2.759)
|
0.071
|
|
Tertile 2
|
Reference
|
|
Reference
|
|
|
Tertile 3
|
3.549 (1.228–9.782)
|
0.014
|
1.708 (0.996–2.926)
|
0.052
|
Abbreviations: CI, confidence interval; HR, hazard ratio.
Note: Age, heart rate (on admission), systolic blood pressure (on admission), creatinine
clearance, sex, diabetes, Killip class (on admission), previous myocardial infarction,
number of diseased vessels, platelet levels on admission were entered into the analysis.
Patients of the T3 group (but not those of the T1 group) had a higher risk of MACE
at 1-year (adjusted HR, 1.530 (1.143–2.049), p = 0.004) and at 5 years of follow-up (adjusted HR: 1.447 (1.1.09–1.887), p = 0.006; [Tables 3] and [Supplementary Table S3] [available in the online version only]) compared with patients of the T2 group (reference
group).
Table 3
The association of platelet count on admission with 1-year and 5-year MACE after adjustment
in the multivariable Cox proportional hazard model
|
1-year MACE
HR (95% CI)
|
p-Value
|
5-year MACE
HR (95% CI)
|
p-Value
|
|
Tertile 1
|
1.199 (0.889–1.618)
|
0.235
|
1.268 (0.973–1.652)
|
0.079
|
|
Tertile 2
|
Reference
|
|
Reference
|
|
|
Tertile 3
|
1.530 (1.143–2.049)
|
0.004
|
1.447 (1.109–1.887)
|
0.006
|
Abbreviations: CI, confidence interval; HR, hazard ratio; MACE, major adverse cardiovascular
events.
Note: Age, heart rate (on admission), systolic blood pressure (on admission), creatinine
clearance, sex, diabetes, Killip class (on admission), previous myocardial infarction,
number of diseased vessels, platelet levels on admission were entered into the analysis.
We also assessed whether high platelet counts are rather a marker of systemic inflammation
induced by large infarcts or whether high platelet counts are causally related to
the increased infarct size and ischemic risk. We tested the collinearity between platelet
counts on admission and other inflammatory markers such as leukocyte counts (on admission
and day 1 after MI) or C-reactive protein (CRP) levels (on admission and day 1 after
MI). Although significant correlations were detected between these variables, Pearson's
correlation coefficients remained <0.3 and variance inflation factors were <6 for
all comparisons ([Supplementary Tables S4–S5] [available in the online version only]) suggesting an insignificant collinearity.
Therefore, leukocyte counts and CRP values were entered into the multivariable Cox
proportional hazard models alongside other baseline variables. The associations between
platelet count on admission and clinical outcomes persisted after including these
markers of systemic inflammation at baseline or on day 1 after MI ([Supplementary Tables S6–S17] [available in the online version only]). These findings suggest that platelet levels
may not only be elevated as part of the inflammatory response induced by large infarcts.
Discussion
Platelets are physiological inhibitors of bleeding from healthy blood vessels (hemostasis)
but can also function as pathological instigators of occlusive events in diseased
blood vessels (thrombosis).[2] MI occurs as a consequence of atherosclerotic plaque rupture/erosion resulting in
atherothrombosis which blocks the supply of oxygenated blood to the heart.[3]
[4] Studies investigating the association of baseline platelet counts with myocardial
infarct size, left ventricular function, and clinical outcome in STEMI patients are
scarce. We here report results from a large STEMI cohort that combines (1) scintigraphic
and enzymatic estimates of infarct size and salvage, (2) left ventricular function
at the time of PPCI and 6 months thereafter, and (3) long-term (5 years) clinical
outcomes (death of any cause and MACE). We found that STEMI patients with highest
baseline platelet count showed greatest infarct size, while STEMI patients with both
highest and lowest baseline platelet count showed increased mortality.
Our study showed that patients with higher platelet counts showed the greatest myocardial
area at risk (the ischemic proportion of the myocardium after coronary occlusion)
as assessed by the first round of SPECT imaging. These data suggest that higher platelet
counts may be related to larger ischemic areas and an increased thrombotic burden.
Apart from larger initial ischemic areas of the myocardium, we also found an association
between higher baseline platelet counts and larger final infarct sizes in the repeat
SPECT imaging. These findings were further corroborated by finding associations between
platelet counts and other markers of infarct size (peak CK-MB and cardiac troponin
T). Although platelet count correlated with the extent of myocardial ischemia and
injury, it was not associated with myocardial salvage after PPCI. These findings may
indicate that platelets impact ischemia (which is caused by atherothrombosis) but
not myocardial rescue or recovery (which may be modified by thromboinflammation).
To our knowledge, this study is the first to address the association between platelet
count and myocardial ischemia in patients with STEMI.
Since patients with higher platelet counts had larger infarct sizes, the strongest
reduction in the left ventricular function was observed in this group. Although the
correlations were close to being significant, the threshold of statistical significance
was not achieved. These findings are in line with a previous publication reporting
that patients with STEMI and higher platelet counts had reduced left ventricular systolic
function both on admission and before discharge.[15] Our study found an association between platelet count on admission and mortality,
which may be explained by larger initial areas at risk and infarct sizes, and more
depressed left ventricular function associated with higher platelet counts. Our finding
that higher platelet counts correlate with higher mortality is in line with observations
of Nikolsky et al who found an independent association between a higher baseline platelet
count and risk of death within the first year after PCI in patients with AMI (mainly
STEMI).[9] Accordingly, Iijima et al showed that a higher platelet count was an independent
correlate of 30-day mortality after PCI in patients with coronary artery disease.[10] Furthermore, Ly et al observed that higher platelet counts were associated with
higher rates of adverse clinical outcomes in patients with STEMI at 30 days.[16] Several studies have reported a U-shaped association between platelet counts (thrombocytosis and thrombocytopenia)
and mortality in patients with ACS. Song et al showed that both low and high platelet
counts were associated with an increased risk of all-cause mortality at 2 years in
patients with acute STEMI or NSTEMI.[7] Małyszczak et al reported that both thrombocytopenia and thrombocytosis on admission
in post-ACS patients were associated with a higher 5-year mortality rate.[8] Mueller et al found that patients with unstable angina/NSTEMI and a platelet count
between 181 and 210 × 109/L had reduced mortality compared with patients with lower or higher platelet counts.[11] These findings agree with our results showing increased mortality in patients with
both highest and lowest baseline platelet counts. Unlike in the T3 group, we did not
find any signal for an impact on the ischemic area, infarct size, and LV function
in the T1 group. This, together with the fact that MACE was also not changed in the
T1 group, may lead to the assumption that the increased mortality observed in the
T1 group may be unrelated to the qualifying event (STEMI) and a consequence of another
underlying disorder. While an increased atherothrombotic risk in patients with high
platelet counts may be suggested, a divergent pathomechanism likely contributes to
increased mortality in patients with low platelet counts.
Our study raises the question whether platelet counts directly impact the pathophysiology
of myocardial ischemia/necrosis or whether platelet counts are rather a marker of
systemic inflammation and the inflammatory response in the acute phase of MI. The
fact that the association between the platelet counts on admission and clinical outcome
persisted after the adjustment for traditional markers of systemic inflammation, such
as leukocyte count and CRP, may suggest that platelets indeed play a direct role in
the pathophysiology of ischemia/necrosis. A study by Tucker et al, in which a mild
reduction of platelet counts without modulating platelet function reduced occlusive
thrombogenesis in a non-human primate model, appears to support this hypothesis.[17] Furthermore, our data showed that a higher platelet count was associated with a
higher risk of mortality at 1 year but not at 5 years after PPCI. We speculate that
this may be explained by the fact that platelets seem to more affect acute ischemia
than recovery after MI (i.e., rather immediate than long-term events). However, this
hypothesis requires further exploration.
Limitations
Our study has several limitations. First, this study represents a retrospective analysis
from an observational study and not from a prospective clinical trial dedicated to
investigating the association of platelet numbers and outcome in patients with STEMI.
In this regard, the study findings should be considered as hypothesis-generating.
Second, since our study enrolled patients between 2002 and 2007, antiplatelet therapy,
although similarly used in between the three groups, was somewhat outdated (was restricted
mostly to clopidogrel) and potent P2Y12 inhibitors were not administered. Consequently,
our results may not be transferable to patients who receive contemporary state-of-the-art
treatment including more potent antiplatelet therapy.
Third, only platelet counts were assessed and data on platelet phenotypes or reticulated
platelets were not available. Hence, we cannot exclude the possibility that the activation
state of platelets might have contributed to our findings. Fourth, although recommendations
on the duration of antiplatelet therapy were based on the guidelines, we cannot exclude
that patients with lower platelet counts received shorter antiplatelet therapy in
the ambulatory setting. Fifths, data on bleeding events were not available. This may
have been of particular interest in the low platelet count group. Here, bleeding may
have been promoted by low platelet numbers or may have caused low platelet numbers
due to platelet consumption. Moreover, data on the management of bleeding including
blood cell transfusions (e.g., administration of red blood cell and/or platelet concentrates)
or platelet function in this group are missing.
