Author + information
- Received June 29, 1998
- Revision received October 9, 1998
- Accepted November 16, 1998
- Published online March 1, 1999.
- Kenneth A Ault, MD∗,* (, )
- Christopher P Cannon, MD§,
- Jane Mitchell∗,
- John McCahan∗,
- Russell P Tracy, PhD†,
- William F Novotny, MD‡,
- James D Reimann, PhD‡ and
- Eugene Braunwald, MD§
- ↵*Reprint requests and correspondence: Dr. Kenneth A. Ault, MMCRI, 125 John Roberts Road, Suite #8, South Portland, Maine 04106
This study was designed to determine the magnitude and time course of platelet activation during therapy of acute coronary syndromes with an oral platelet antagonist.
Platelet activation and aggregation are central to the pathogenesis of the acute coronary syndromes (ACS). However, few data are available on levels of platelet activation over time in patients with ACS, especially in the setting of chronic glycoprotein (GP) IIb/IIIa inhibition.
The Thrombolysis in Myocardial Infarction (TIMI) 12 trial was a phase II, double-blind trial evaluating the effects of sibrafiban, an oral, selective antagonist of the platelet glycoprotein IIb/IIIa receptor in patients stabilized after an ACS. A subset of 90 of the 329 patients in the study had measurement of platelet activation as assessed by the expression of platelet associated P-Selectin on days 0, 7 and 28. Platelet activation was measured in blood samples that were fixed either immediately (spontaneous activation) or after 5 minute incubation with 0, 1 μM or 5 μM ADP in order to assess platelet responsiveness to very low or moderate stimulation.
At baseline there was a significant elevation of spontaneous platelet activation as compared to samples obtained from normal donors or from patients who did not have acute coronary syndromes (ACS patients 27.6 ± 18.7%, Normal controls 8.5 ± 4.4%, Patient controls 10.9 ± 7.1%, p < 0.005 for both). In addition, there was a significant decrease in the levels of platelet activation with time during the 28 days of treatment with sibrafiban. Nevertheless, even on day 28, the TIMI-12 patients continued to show elevated platelet activation in comparison to the control groups (p < 0.05 for both).
These results suggest that platelets remain activated long after clinical stabilization post ACS. Although platelet activation decreased after one month of oral GPIIb/IIIa inhibition, levels remained higher than normal, suggesting the need for long-term antiplatelet therapy following ACS.
There is abundant evidence that platelet activation occurs during acute coronary syndromes and that measurement of platelet activation and/or aggregation, may provide prognostic information in patients suffering a first or second event (1–11). In addition, there is evidence that inhibition of platelet function through the use of aspirin and inhibitors of platelet glycoprotein (GP) IIb/IIIa function lowers the risk of first and subsequent cardiac events (12–16). What remains uncertain is to what extent the increased platelet activation observed is due to the disease process itself, i.e., activation of platelets within the abnormal coronary circulation, or to a genetic or environmental propensity for platelets to become activated more easily in response to minimal stimuli. In the former case, one would expect that platelet status would return to normal early after the resolution of the acute event, whereas, in the latter case platelet function may remain abnormal for a considerable time or indefinitely. With the advent of potentially much more powerful platelet inhibitors such as oral GPIIb/IIIa antagonists (17–19), the possibility of prolonged treatment with these drugs may make it important to determine both the extent and the duration of platelet dysfunction following a coronary event.
We (20–24)and others (1,7,25–27)have used platelet associated P-Selectin as a sensitive, quantitative measure of platelet activation. The P-Selectin molecule has the advantage that its antigenic epitopes are relatively stable to fixation and it thus lends itself to multicenter clinical studies. The TIMI-12 trial was a study of the safety and effectiveness of prolonged (28 day) administration of an oral glycoprotein IIb/IIIa inhibitor (sibrafiban) in patients who had suffered an acute coronary syndrome (18). P-Selectin measurements were included in the pharmacokinetic/pharmacodynamic cohort of the TIMI-12 trial in order to determine if there were detectable changes in platelet activation during the course of the disease process and whether the study drug had an effect on platelet activation in addition to its known effect on platelet aggregation.
Materials and methods
The monoclonal antibody 1E3, a CD62P antibody directed against P-Selectin was derived in our laboratory as previously reported (23). It was biotinylated by standard methods. Phycoerythrin conjugated streptavidin was obtained from Southern Biotech, Birmingham, Alabama. Fluorescein conjugated monoclonal antibody directed against glycoprotein GPIIb/IIIa (CD41) was obtained from Immunotech Inc., Westbrook, Maine. All reagents were used in saturating concentrations as determined by titration on fully activated platelet samples.
A complete description of the clinical characteristics of the TIMI-12 subjects and the sites participating has been previously published (18). Briefly, the subjects had suffered the onset of an acute coronary syndrome (unstable angina, non-Q-wave myocardial infarction [MI] or Q-wave MI) between 24 hours and 7 days prior to randomization. Patients were required to be clinically stable, i.e., no ongoing ischemia, no congestive heart failure, and not on intravenous heparin. They were randomized to receive one of seven different dose regimens of sibrafiban for 28 days; all subjects received some dose of study drug in a double-blind fashion. No other GPIIb/IIIa antagonists were used. Platelet activation studies were done on 90 patients (41 with unstable angina, 30 with non-Q-wave MI and 19 with Q-wave MI) at enrollment (day 0) and on days 7 and 28. In addition to the TIMI-12 subjects this report also includes data obtained from two other groups of subjects. One consisting of normal subjects and the other consisting of patients studied in the Maine Medical Center emergency department who were being admitted for various acute medical problems not involving cardiovascular disease. The mean age of the groups were: normal subjects, 36 ± 10 (sd), patient controls 66 ± 13.3, TIMI-12 patients 57 ± 11. Fifty-five percent of the normal subjects, 40% of the patient controls and 77% of the TIMI-12 patients were male. All subjects were studied under protocols approved by the appropriate institutional review boards for human research.
Blood samples were obtained through clean venipuncture or free flowing intravenous lines after discarding the void volume. All blood samples were collected in heparinized tubes (Becton-Dickinson Vacutainer, Franklin Lakes, New Jersey) and subsequent processing was done immediately at the bedside using kits containing all of the required reagents and supplied to the participating centers by our laboratory. One drop (50 μL) of freshly drawn blood was placed immediately into a tube containing 1 mL of 2% formaldehyde in phosphate buffered saline (PBS) solution and mixed. Three aliquots of 0.5 mL each were placed into three other tubes each containing zero, 1 μM or 5 μM (final concentration) of adenosine diphosphate (ADP). After mixing, the tubes were incubated for 5 minutes at room temperature, and then 50 μL from each tube was transferred into three additional tubes containing 1 mL of 2% formaldehyde. After 30 minutes at room temperature, the fixed samples were diluted to 5 mL with PBS to dilute the fixative. The result was four tubes of fixed whole blood, one had been fixed immediately (spontaneous activation) and the other three represented the result of activation after a 5 minute exposure to 0, 1 μM, and 5 μM ADP.
The fixed samples were refrigerated and shipped overnight to our laboratory. The samples from the control subjects were obtained and processed in exactly the same way, using the same kits and the same reagents. On arrival, the samples were washed with PBS to remove the fixative and then labeled with saturating amounts of fluorescein conjugated CD41 and biotinylated CD62P monoclonal antibodies. After 20 minutes incubation at room temperature, the samples were washed once and incubated with phycoerythrin conjugated streptavidin to label the bound biotinylated antibody. After a further 20-minute incubation the samples were washed again and analyzed using a Becton Dickinson FACScan flow cytometer (Becton-Dickinson Immunocytometry Systems, San Jose, California). The details of this analysis have been previously published (21,22). Briefly the platelets were identified using the CD41 monoclonal which labeled all platelets, and the percentage of P-Selectin positive platelets was determined by reference to a negative control from which the CD62P monoclonal had been omitted. Ten thousand platelets were analyzed in each sample. In those samples that were stimulated with ADP, there was often significant aggregation. The aggregates, which also express elevated P-Selectin, were included in the count of activated platelets. In the samples corresponding to spontaneous and zero ADP stimulation, there was no significant aggregation and thus the measure of percent activated platelets was not affected.
Statistical comparisons of baseline with day 7 and day 28 results within the TIMI-12 data were done using least squares means of the change from baseline. Comparisons of baseline platelet activation with the control groups were done using unpaired Student’s ttest. The significance of the relationship between platelet activation and index event or initial therapy was tested by two way repeated measures ANOVA after checking that the interaction between baseline characteristics and ADP dose was not significant.
Changes in platelet activation with time in the TIMI-12 subjects
Spontaneous platelet activation was significantly elevated at baseline (27.6 ± 18.7%) and decreased with time over the 28 day treatment period to 16.7 ± 13.4% at day 7 (p < 0.001) and 20.0 ± 13.7% at day 28 (p = 0.021). Similar decreases were seen for the response to zero and 1 μM ADP but not for the response to 5 μM ADP (Fig. 1). All measures of platelet activation at baseline on day 7 and on day 28 showed no correlation with the dose of study drug (data not shown). For this reason, the data for subjects treated at all seven drug doses were combined for subsequent analysis.
Comparison of TIMI-12 results with control groups
When compared to either control group, the levels of both spontaneous platelet activation, and the response to no ADP were significantly higher not only at baseline but also at day 28 (Table 1). The responses to 1 μM and 5 μM ADP were not significantly different from those of either control group. The two control groups were not significantly different from each other.
Platelet activation and baseline covariates in the TIMI-12 patients
Platelet activation response was significantly correlated with the type of index event (Fig. 2, upper panel). Higher levels of platelet responses were seen in those patients having Q-wave MI (19 patients), less in those with non-Q-wave MI (30 patients) and least in those with unstable angina (41 patients). Similarly, platelet activation was significantly correlated with the type of treatment used for the index event, prior to obtaining the baseline sample (Fig. 2, lower panel). Patients treated with thrombolysis (18 patients) and primary percutaneous transluminal coronary angioplasty (PTCA) (15 patients) had higher levels of activation than those with neither intervention (57 patients). When analyzed separately, it was clear that most of the elevated baseline platelet activation was in those patients who received either thrombolysis or PTCA as primary therapy. Nevertheless, those receiving neither intervention also had a decrease in platelet activation over time which was significant for the 0 μM ADP response but not for spontaneous activation (Table 2). There were no significant correlations between any measure of platelet activation and age, weight, race or gender (data not shown).
Platelet activation and other laboratory parameters
Weak correlations were observed between spontaneous platelet activation and acute phase reactants c-reactive protein and fibrinogen levels (r values of 0.2 and 0.3, p values 0.02 and 0.04, respectively). In addition, there was a weak correlation between all measures of platelet activation and D-dimer levels (r values between 0.2 and 0.35, p values between 0.004 and 0.05). No significant correlations were observed with fibrinopeptide A or prothrombin activation fragment 1 + 2.
Our results strongly suggest that patients presenting with acute coronary syndromes have increased levels of platelet activation that decrease with time after the event. The design of this study does not allow us to clearly determine whether the decreased platelet activation was due to the drug treatment or to overall resolution of the coronary lesion over time, although the lack of any correlation with drug dose makes an effect of the drug less likely. The comparison of the patients in TIMI-12 with two different control groups also suggests that elevated platelet activation persists for a considerable period in many patients following acute coronary syndromes (ACS).
Although baseline platelet activation was higher for those receiving treatment with either thrombolysis or PTCA, there was also a decrease over time in those patients receiving neither intervention. This suggests that increased platelet activation may be due to both therapy and the underlying disease process.
We have observed weak correlations between platelet activation parameters and some acute phase reactants such as C-reactive protein and fibrinogen which also decrease over time following the acute event. Although these correlations are not surprising, the weakness of the correlations suggests that measurement of platelet activation is providing a different type of information about the disease process than these other measures of acute inflammation.
One question that must be addressed is whether the relatively high levels of platelet activation we observed in the specimens that were fixed immediately (spontaneous activation) are actually present in vivo. When blood samples are obtained from normal subjects under carefully controlled laboratory conditions, labeled and analyzed immediately, with or without fixation, considerably lower levels of platelet activation are observed, in the range of 1 to 5% (2,28). Thus, we feel that the levels of activation reported here for normals are due to the fact that these samples were obtained under “clinical conditions” and were not analyzed immediately. The patient and control samples were all obtained and the analyses performed under the same conditions, thus permitting valid comparison between the groups. Some of the activation observed even in the immediately fixed samples may have occurred during or after the drawing of the sample, prior to fixation. Further studies will be necessary to make this distinction. Although the normal subjects were significantly younger than the patient control group, their platelet activation was not different. This confirms our previous experience that platelet activation does not change dramatically with age (unpublished data).
The fact that these results could be obtained in a multicenter study using “kits” sent out from a central laboratory suggests that the measurement of platelet associated P-Selectin is a relatively robust test that may find clinical application in the future. It has been suggested that measurement of soluble P-Selectin (29–31)may also reflect increased platelet activation in similar settings. Since soluble P-Selectin could in theory arise either from platelets or from endothelial cells, the question of whether platelet associated P-Selectin and soluble P-Selectin are providing the same or different information about the disease process remains to be answered.
Chronic oral GPIIb/IIIa inhibition
With the advent of both parenteral and oral agents that strongly inhibit platelet function, physicians now have tools that permit direct manipulation of platelet function during and after a coronary event. The questions that remain to be answered are which patients should receive these new agents and for how long they should be administered in order to prevent platelets from triggering or participating in subsequent thrombosis. Our results suggest that measurement of platelet activation may be one way to determine which patients have a major component of platelet involvement in their disease process and thus may benefit most from platelet inhibition. They also suggest that prolonged inhibition of platelet function may be needed to allow platelets to return to baseline low levels of responsiveness, thereby reducing the risk of recurrent ischemic events. This hypothesis for the potential clinical benefit of long-term oral GPIIb/IIIa inhibition is currently being tested in large clinical trials.
☆ This work was supported by Genentech Inc., South San Francisco, CA, and by Hoffman La Roche, Ltd. Basel, Switzerland.
- acute coronary syndromes
- adenosine diphosphate
- myocardial infarction
- percutaneous transluminal coronary angioplasty
- phosphate buffered saline
- Thrombolysis in Myocardial Infarction
- Received June 29, 1998.
- Revision received October 9, 1998.
- Accepted November 16, 1998.
- American College of Cardiology
- Abrams C,
- Shattil S.J
- Kolarov P,
- Tschoepe D,
- Nieuwenhuis H.K,
- Gries F.A,
- Strauer B,
- Schultheiss H.P
- Michelson A.D
- Nurden A.T,
- Macchi L,
- Bihour C,
- Durrieu C,
- Besse P,
- Nurden P
- Patrono C,
- Renda G
- The RESTORE Investigators
- The Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrelin Therapy (PURSUIT) Trial Investigators
- Ferguson J,
- Deedwania P,
- Kereiakas D,
- et al.
- Cannon C,
- McCabe C,
- Borzak S,
- et al.
- Kereiakes D,
- Runyon J,
- Kleiman N,
- Higby N,
- Anderson L,
- Hantsbarger G,
- McDonald S,
- Anders R
- Abrams C.S,
- Ellison N,
- Brudzynski A.Z,
- Shattil S.J
- Shattil S.J,
- Cunningham M,
- Hoxie J.A
- McEver R.P,
- Martin M.N
- Ikeda H,
- Takajo Y,
- Ichiki K,
- et al.