- Email: [email protected]
IIIa receptor inhibitors: A pooled analysis
Thrombocytopenia complicating treatment with intravenous glycoprotein IIb/IIIa receptor inhibitors: A pooled analysis Himadri Dasgupta, MD, James C. Blankenship, MD, G. Craig Wood, MS, Carolin M. Frey, PhD, S. Lee Demko, RN, and Francis J. Menapace, MD Danville, Pa
Background Despite the increasingly prevalent role of platelet glycoprotein (GP) IIb/IIIa receptor inhibitors in acute coronary syndromes and percutaneous coronary interventions, the incidence and clinical relevance of thrombocytopenia occurring with their use remain unclear.
Methods We identified 8 placebo-controlled, randomized, large trials of GP IIb/IIIa receptor inhibitors reporting the incidence of thrombocytopenia, grouped by severity. The clinical courses of 42 patients with GP IIb/IIIa-related thrombocytopenia in these studies and other case reports were reviewed for bleeding complications.
Results Abciximab increased mild thrombocytopenia compared with placebo (4.2% vs 2.0%; P < .001; odds ratio 2.14) and increased severe thrombocytopenia compared with placebo (1.0% vs 0.4%; P = .01; odds ratio 2.48). Smallmolecule IIb/IIIa inhibitors did not significantly increase mild or severe thrombocytopenia compared with placebo. Mild thrombocytopenia occurred more frequently in acute coronary syndrome trials than in coronary intervention trials, even in patients not receiving any IIb/IIIa inhibitors. No major bleeding sequelae were reported in 23 patients with severe thrombocytopenia or in 19 patients with profound thrombocytopenia.
Conclusions Abciximab, but not eptifibatide or tirofiban, increases the incidence of thrombocytopenia compared with placebo in patients also treated with heparin. Thrombocytopenia associated with GP IIb/IIIa inhibition does not routinely lead to severe bleeding complications. (Am Heart J 2000;140:206-11.)
Platelets play a central role in initiating and mediating ischemia and related complications during acute coronary syndromes (ACS) and percutaneous coronary interventions (PCI). Platelet glycoprotein (GP) IIb/IIIa receptor inhibitors block the final common pathway to platelet aggregation and have played an increasing role in the management of ACS and PCI. However, these agents may cause internal bleeding and bleeding at the sites of vascular injury. Thrombocytopenia occurring with GP IIb/IIIa inhibitors has not been well-characterized. The relative propensity of various GP IIb/IIIa inhibitors to cause thrombocytopenia may affect their acceptance in clinical practice. We determined the incidence of thrombocytopenia occurring with different GP IIb/IIIa receptor inhibitors and its relation to clinically significant bleeding. From Geisinger Medical Center, Penn State Geisinger Health System, and The Janet Weis Research Center. Presented in part at the American College of Cardiology 48th Annual Scientific Sessions, March 9, 1999, New Orleans, La. Submitted December 27, 1999; accepted March 29, 2000. Reprint requests: James C. Blankenship, MD, Department of Cardiology 21-60, Geisinger Medical Center, Danville, PA 17822. E-mail: [email protected] Copyright © 2000 by Mosby, Inc. 0002-8703/2000/$12.00 + 0 4/1/107554 doi:10.1067/mhj.2000.107554
Methods Patients All major prospective, placebo-controlled, randomized, multicenter trials of intravenous GP IIb/IIIa receptor inhibitors were identified by a MEDLINE search. Only trials that clearly defined and documented levels of thrombocytopenia were included (Table I).1-8 Multicenter trials of lamifiban were excluded because they did not report specific levels or incidence of thrombocytopenia. Several trials1,4,5,8 excluded patients with baseline thrombocytopenia platelet count below 90,000 to 100,000/µL; others excluded patients with hemorrhagic diathesis but did not list specific exclusion criteria based on baseline platelet counts.2,3,6,7 Reports of large IIb/IIIa inhibitor trials were reviewed to identify descriptions of patients with severe thrombocytopenia, and a MEDLINE search was used to identify case reports or small studies of thrombocytopenia induced by GP IIb/IIIa inhibitors.
Definitions of thrombocytopenia Definitions of various levels of thrombocytopenia are not uniform. However, most studies define mild thrombocytopenia as a platelet count of <90,000 to 100,000/µL, severe thrombocytopenia <50,000/µL, and profound thrombocytopenia as <20,000/µL. The incidence of mild and severe thrombocytopenia in the selected studies was identified for each type of IIb/IIIa receptor inhibitor (abciximab, eptifibatide, and tirofiban) (Table I).
American Heart Journal Volume 140, Number 2
Statistical analysis The conditional test based on the odds ratio (OR) was used to compare the incidence of thrombocytopenia in patients treated with IIb/IIIa receptor inhibitor versus placebo.9 StatExact version 3.1 (Cytel Software, Cambridge, Mass) was used to calculate exact confidence intervals and P values. Additional analyses with these statistical tests aggregated trials by different characteristics (eg, ACS vs PCI, abciximab vs small-molecule drugs) to determine whether these differences in trial design affected the incidence of thrombocytopenia in patients receiving placebo/heparin and in patients receiving IIb/IIIa inhibitor/heparin. Statistical comparisons were not made across aggregated groups (eg, incidence of thrombocytopenia in heparin/placebo groups in ACS trials vs heparin/placebo groups in PCI trials). We avoided such statistical comparisons because patients admitted to one trial may have differed systematically from those enrolled in another regarding factors that might affect their risk of thrombocytopenia. Because of the multiple comparisons made among groups, conclusions about statistical significance were made only with extreme caution.
Results Eight studies that fit the inclusion criteria were identified (Table I). Table II shows the incidence of mild thrombocytopenia in patients receiving specific types of IIb/IIIa receptor inhibitor with heparin versus placebo with heparin. Abciximab increased the incidence of mild thrombocytopenia compared with the placebo group (4.2% vs 2.0%; P < .001; OR 2.13). Eptifibatide or tirofiban with heparin did not increase mild thrombocytopenia compared with placebo with heparin (OR 0.99). Table III shows the incidence of severe thrombocytopenia in the same groups as in Table II. Patients receiving abciximab with heparin had more than twice the incidence of severe thrombocytopenia than those receiving placebo with heparin (1.0% vs 0.4%; P = .01; OR 2.48). Eptifibatide or tirofiban with heparin did not cause a significant excess of severe thrombocytopenia compared with placebo with heparin (0.3% vs 0.2%, P = .16). In the above analyses, abciximab data came predominantly from PCI trials, and small-molecule drug data came predominantly from ACS trials. To determine if the lack of thrombocytopenia with small-molecule inhibitors was obscured by the design of ACS trials, we compared small-molecule inhibitor ACS trials with small-molecule inhibitor PCI trials (Table IV). Smallmolecule IIb/IIIa inhibitors did not cause a significant excess of thrombocytopenia in either coronary intervention trials or ACS trials. (This analysis could not be performed with abciximab because no data were available from abciximab ACS trials.) To exclude type of trial (ACS vs PCI) as a possible confounding factor, these analyses were repeated excluding all ACS trials (Table V). PCI trials were aggregated by type of IIb/IIIa inhibitor (abciximab vs smallmolecule inhibitors). As before, abciximab, but not
Dasgupta et al 207
the small-molecule inhibitors, was associated with an excess of mild and severe thrombocytopenia compared with placebo. Table II shows that patients receiving placebo (not receiving any IIb/IIIa inhibitor) developed mild thrombocytopenia in small-molecule IIb/IIIa trials (4.6%) more often than in abciximab trials (2.0%). This variation in the incidence of thrombocytopenia among placebo groups from different trials raised the possibility that trial design affected the incidence of thrombocytopenia. To detect such an effect, trials were aggregated by type of trial (ACS vs PCI) rather than by type of IIb/IIIa inhibitor (Table VI). Patients receiving placebo had mild thrombocytopenia develop in 5.8% of ACS trials versus 1.9% in PCI trials. This analysis was repeated excluding abciximab trials; when small-molecule IIb/IIIa inhibitor ACS trials were compared with small-molecule IIb/IIIa inhibitor PCI trials (Table IV), placebo groups in ACS trials had 3-fold more mild thrombocytopenia than placebo groups in PCI trials (5.8% vs 1.7%). Table V provides additional data indicating that differences in incidence of thrombocytopenia in placebo groups among studies were caused by the type of trial (ACS vs PCI), not type of IIb/IIIa inhibitor used. When ACS trials were excluded, abciximab PCI trials and small-molecule IIb/IIIa PCI trials produced very similar incidences of mild thrombocytopenia in patients receiving placebo with heparin (2.0% vs 1.7%). The outcomes of 42 patients, 23 with severe thrombocytopenia and 19 with profound thrombocytopenia, are listed in Table VII.4,5,10-18 None of these involved retreatment with abciximab. No major bleeding complication was reported in any of the patients despite levels of thrombocytopenia as low as <10,000/µL. Thrombocytopenia was transient and resolved either spontaneously or with platelet transfusions.
Discussion The most important finding of our analysis is the significant increase in thrombocytopenia (mild and severe) when used with abciximab and heparin compared with placebo and heparin. Although there was a trend toward a similar increase in severe thrombocytopenia with other GP IIb/IIIa receptor inhibitors, it was not statistically significant. Patients treated with placebo had higher incidence of mild thrombocytopenia in ACS studies (compared with PCI trials), suggesting that the design of these studies may have increased the incidence of thrombocytopenia regardless of the IIb/IIIa inhibitor used in the trial. This trend was not observed for severe thrombocytopenia. It is likely that the longer infusions of heparin used in ACS trials account for more frequent mild thrombocytopenia. Heparin-induced thrombocytopenia rarely causes profound thrombocytopenia. Consistent with that, there was no increase in profound thrombocytopenia
American Heart Journal August 2000
208 Dasgupta et al
Table I. Studies reporting thrombocytopenia with GP IIb/IIIa receptor inhibitors
Study CAPTURE1 EPIC2 EPILOG3 PRISM4 PRISM-PLUS5 PURSUIT6 IMPACT II7 RESTORE8
Study drug
Patient population
Placebo + heparin (n)
GP IIb/IIIa + heparin (n)
A A A T T E E T
PCI PCI PCI ACS ACS ACS PCI PCI
635 696 939 1616 797 4696 1285 1070
630 1403 1853 — 773 4679 2586 1071
GP IIb/IIIa (n) — — — 1616 345 — — —
Empty cells reflect drug combinations not used in the trials. A, Abciximab, T, tirofiban, E, eptifibatide.
Table II. Mild thrombocytopenia according to drug groups Treatment group Heparin + any GP IIb/IIIa Heparin + abciximab Heparin + small-molecule GP IIb/IIIa GP IIb/IIIa alone‡
Incidence in GP IIb/IIIa + heparin groups
n* 7 3 4 1
4.5% 4.2% 4.6% 1.1%†
Incidence in placebo + heparin groups
OR
95% confidence interval for OR
Exact P value
4.1% 2.0% 4.6% 0.4%†
1.11 2.13 0.99 3.02
(0.96-1.25) (1.52-3.04) (0.86-1.15) (1.15-9.32)
.17 <.001 .94 .02
Incidence in placebo + heparin groups
OR
95% confidence interval for OR
Exact P value
0.2% 0.4% 0.2% 0.1%
2.31 2.48 1.73 3.01
*Number of studies contributing to comparison. †These patient groups received GP IIb/IIIa inhibitors without heparin. ‡PRISM (using tirofiban in patients with ACS).
Table III. Severe thrombocytopenia according to drug groups Treatment group Heparin + any GP IIb/IIIa Heparin + abciximab Heparin + small-molecule GP IIb/IIIa GP IIb/IIIa alone†
n* 7 3 4 1
Incidence in GP IIb/IIIa + heparin groups 0.5% 1.0% 0.3% 0.4%*
(1.38-3.98) (1.18-5.85) (0.83-3.79) (0.54-30.50)
<.001 .01 .16 .29
These patient groups received GP IIb/IIIa inhibitors without heparin. *Number of studies contributing to comparison. †PRISM (using tirofiban in patients with ACS).
in ACS versus PCI trials. An increase in heparin-induced thrombocytopenia from the longer infusions used in ACS trials (versus PCI trials) might explain in part the higher incidence of thrombocytopenia in patients receiving IIb/IIIa inhibitors in ACS trials (vs PCI trials). Our study documents the relatively benign clinical course of thrombocytopenia associated with first exposure to GP IIb/IIIa receptor inhibitors. Access site hematomas and minor oozing were the most frequent
bleeding complications. One patient had hematemesis without other complications.15 Thrombocytopenia was transient and either improved spontaneously or with platelet transfusions (in one instance with additional intravenous immunoglobulin and steroids).10 The cause for the observed difference in incidence of thrombocytopenia induced by abciximab as opposed to eptifibatide or tirofiban is unclear. Abciximab induces an antibody response (human antichimeric antibod-
American Heart Journal Volume 140, Number 2
Dasgupta et al 209
Platelet count <90,000-100,000/µL Placebo + heparin (n)
GP IIb/IIIa + heparin (n)
8 24 14 6 6 315 31 10
<50,000/µL GP IIb/IIIa (n)
35 82 47 — 15 318 71 13
Placebo + heparin (n)
GP IIb/IIIa + heparin (n)
GP IIb/IIIa (n)
0 5 4 2 2 4 5 1
12 14 12 — 4 9 9 2
— — — 6 — — — —
— — — 18 — — — —
Table IV. Thrombocytopenia in small-molecule IIb/IIIa trials aggregated by ACS versus PCI trials Platelet count (per µL)
Type of trial
No. of studies
No. of patients
<90,000-100,000 <90,000-100,000 <50,000 <50,000
ACS PCI ACS PCI
2 2 2 2
10,945 6012 10,945 6012
Incidence in GP IIb/IIIa + heparin groups 6.1% 2.3% 0.2% 0.3%
Incidence in placebo + heparin groups 5.8% 1.7% 0.1% 0.3%
OR
95% Confidence interval for OR
Exact P value
1.05 1.33 2.19 1.18
(0.89-1.23) (0.90-1.99) (0.77-7.02) (0.40-3.90)
.59 .16 .16 .95
Exact P value .16 <.001 .95 .01
Table V. Thrombocytopenia in PCI trials by type of IIb/IIIa inhibitor Platelet count (per µL)
Type of drug
No. of studies
No. of patients
Incidence in GP IIb/IIIa + heparin groups
Incidence in placebo + heparin groups
OR
95% Confidence interval for OR
<90,000-100,000 <90,000-100,000 <50,000 <50,000
Small-molecule Abciximab Small-molecule Abciximab
2 3 2 3
6012 6156 6012 6156
2.3% 4.2% 0.3% 1.0%
1.7% 2.0% 0.3% 0.4%
1.33 2.13 1.18 2.48
(0.90-1.99) (1.52-3.04) (0.40-3.90) (1.18-5.85)
ies).19,20 These antibodies may lead to increased platelet consumption and contribute to higher incidence of thrombocytopenia with abciximab. Other mechanisms of thrombocytopenia with abciximab may include the development and subsequent removal of aggregates of platelets and c7E3 Fab, binding of preexisting immunoglobulin G antibodies to abciximab-induced new platelet epitopes (ligand-induced binding sites), and other kinds of platelet modifications resulting from interaction with abciximab.21-24 The contribution, if any, of these mechanisms to the observed difference
in thrombocytopenia caused by abciximab versus other GP IIb/IIIa receptor inhibitors is currently unknown. Limitations of this study include those of any metaanalysis. It is retrospective, comparing trials that differ to some extent in drugs, treatment protocols, monitoring protocols, and definitions. Several large trials of GP IIb/IIIa receptor inhibitors could not be included because they did not report the incidence of thrombocytopenia. Because of the multiple comparisons made among groups, conclusions about statistical significance should be drawn very cautiously. Ideally, the rel-
American Heart Journal August 2000
210 Dasgupta et al
Table VI. Thrombocytopenia: ACS versus PCI trials Platelet count (per µL)
Type of trial
No. of studies
No. of patients
Incidence in GP IIb/IIIa + heparin groups
Incidence in placebo + heparin groups
<90,000-100,000 <90,000-100,000 <50,000 <50,000
ACS PCI ACS PCI
2 5 2 5
10,945 12,168 10,945 12,168
6.1% 3.3% 0.2% 0.6%
5.8% 1.9% 0.1% 0.3%
OR
95% Confidence interval for OR
Exact P value
1.05 1.77 2.19 2.01
(0.89-1.23) (1.38-2.30) (0.77-7.02) (1.11-3.86)
.59 <.001 .16 .02
Table VII. Clinical consequences of GP IIb/IIIa-induced severe and profound thrombocytopenia Study PRISM4 PRISM-PLUS5 EPIC10 CAPTURE1 Kereiakes et al11
Patients with platelet count <50,000/µL 6 4
2 10
Berkowitz et al12 Mendes et al13 Jenkins et al14 Jubelirer et al15 Ferrari et al16 Joseph et al17 Elmi et al18
Patients with platelet count <20,000/µL
2 4 4 1 1 3
3 1 1
ative contribution of IIb/IIIa inhibitors and heparin to thrombocytopenia could be assessed by multiple logistic regression analysis of the original trial databases pooled together, but without access to these databases this approach is impossible. The individual reports of GP IIb/IIIa-associated thrombocytopenia do not allow comparison of bleeding sequelae in patients who did not receive these drugs. However incomplete, these cases suggest that severe bleeding is unusual (none of 42 patients) with GP IIb/IIIa-associated thrombocytopenia, and it highlights the need for definition and reporting of thrombocytopenia in large GP IIb/IIIa trials. In summary, mild and severe thrombocytopenia occur more frequently in patients treated with abciximab and heparin than with placebo and heparin. Tirofiban and eptifibatide alone or in combination with heparin did not result in a statistically significant higher incidence of thrombocytopenia than heparin use alone. ACS trials tended to report higher incidences of mild thrombocytopenia (compared with coronary intervention trials), perhaps because of longer heparin infusions producing heparin-induced thrombocytopenia. Severe and profound thrombocytopenia from GP IIb/IIIa receptor inhibitors
Bleeding complications and interventions No clinical sequelae No sequelae No bleeding; platelet transfusion in both, intravenous immunoglobulin + steroid in one No bleeding; 5 received platelet transfusions 3 Groin, 1 forearm hematoma, no major bleeding; all received platelet transfusions, 2 received 2 U red blood cell transfusion each No major bleeding; small oozing + groin hematoma in 1 No bleeding; received platelet transfusion No bleeding Hematoma of groin, forearm, and scrotum; hematemesis in 1 (platelet and blood transfusion [2 U] in this patient only) No bleeding No bleeding No bleeding, prophylactic platelet transfusion given
does not often cause severe bleeding. Future trials should clearly define and report mild, severe, and profound thrombocytopenia (platelet counts <100,000/µL, <50,000/µL, and <20,000/µL, respectively) to facilitate comparison of GP IIb/IIIa receptor inhibitors. Future trials should also evaluate patients with thrombocytopenia for heparin-induced thrombocytopenia.
References 1. The CAPTURE investigators. Randomized placebo-controlled trial of abciximab before and during coronary intervention in refractory unstable angina: the CAPTURE study. Lancet 1997;349:1429-35. 2. The EPIC investigators. Use of a monoclonal antibody directed against the platelet glycoprotein IIb/IIIa receptor in high risk coronary angioplasty. N Engl J Med 1994;330:956-61. 3. The EPILOG investigators. Platelet glycoprotein IIb/IIIa receptor blockade and low dose heparin duration percutaneous coronary revascularization. N Engl J Med 1997;336:1689-96. 4. The PRISM study investigators. A comparison of aspirin plus tirofiban with aspirin plus heparin for unstable angina. N Engl J Med 1998; 338:1498-505. 5. The PRISM-PLUS study investigators. Inhibition of the platelet glycoprotein IIb/IIIa receptor with tirofiban in unstable angina and nonQ-wave myocardial infarction. N Engl J Med 1998;338:1488-97.
American Heart Journal Volume 140, Number 2
6. The PURSUIT trial investigators. Inhibition of platelet glycoprotein IIb/IIIa with eptifibatide in patients with acute coronary syndromes. N Engl J Med 1998;339:436-43. 7. The IMPACT-II investigators. Randomized placebo- controlled trial of eptifibatide on complications on percutaneous coronary interventions: IMPACT-II. Lancet 1997;347:1422-8. 8. The RESTORE investigators. Effect of platelet glycoprotein IIb/IIIa blockade with tirofiban on adverse cardiac events in patients with unstable angina or acute myocardial infarction undergoing coronary angioplasty. Circulation 1997;96:1445-53. 9. Gart J. Point and interval estimation of the common odds ratio in the combination of 2X2 tables with fixed margins. Biometrika 1970; 57:471-5. 10. The EPIC study group. Occurrence and clinical significance of thrombocytopenia in a population under going high-risk percutaneous revascularization. J Am Coll Cardiol 1998;32:311-9. 11. Kereiakes DJ, Essell JH, Abbottsmith CW, et al. Abciximab-associated profound thrombocytopenia: therapy with immunoglobulin and platelet transfusion. Am J Cardiol 1996;78:1161-3. 12. Berkowitz SD, Harrington RA, Rund MM, et al. Acute profound thrombocytopenia after c7E3 Fab (abciximab) therapy. Circulation 1997;95:809-13. 13. Mendes V, Resende M, Santos JM, et al. Trombocitopenia grave associada a terapeutica com abciximab-a proposito de um caso clinico. Revista Portuguesa de Cardiologia 1998;17:609-13. 14. Jenkins LA, Lau S, Crawford M, et al. Delayed profound thrombocytopenia after c7E3 Fab (abciximab) therapy [letter]. Circulation 1998;97:1214-5. 15. Jubelirer SJ, Koenig BA, Bates MC. Acute profound thrombocytopenia following c7E3 Fab (abciximab) therapy: case reports, review of literature and implications for therapy. Am J Hematol 1999;61:205-8.
Dasgupta et al 211
16. Ferrari E, Thiry M, Touati C, et al. Acute profound thrombocytopenia after c7E3 Fab therapy [letter]. Circulation 1997;95: 3809-10. 17. Joseph T, Marco J, Grogorini L. Acute profound thrombocytopenia after abciximab therapy during coronary angioplasty. Clin Cardiol 1998;21:851-2. 18. Elmi F, Oza R, Mascarenhas DAN. Acute profound thrombocytopenia without bleeding complications after abciximab administration. J Invas Cardiol 1999;11:313-5. 19. Berkowitz SD, Sane DC, Sigmon KN, et al. Occurrence and clinical significance of thrombocytopenia in a population undergoing highrisk percutaneous coronary revascularization. J Am Coll Cardiol 1998;32:311-9. 20. Jordan RE, Wagner CL, Mascelli MA, et al. Preclinical development of c7E3 Fab: a mouse/human chimeric monoclonal antibody fragment that inhibits platelet function by blockade of GP IIb/IIIa receptors with observations on the immunogenicity of c7E3 Fab in humans. In: Horton MA, editor. Adhesion receptors as therapeutic targets. New York: CRC Press; 1996. p 281-302. 21. Frelinger AL III, Cohen I, Plow EF, et al. Selective inhibition of integrin function by antibodies specific for ligand-occupied receptor conformers. J Biol Chem 1990;265:6346-52. 22. Bednar B, Bednar RA, Cook JJ, et al. Drug-dependent antibodies against GPIIb/IIIa induce thrombocytopenia [abstract]. Circulation 1996;94 suppl I:I-99. 23. Coller BS, Scudder LE, Beer J, et al. Monoclonal antibodies to platelet glycoprotein IIb/IIIa as antithrombotic agents. Ann NY Acad Sci 1991;614:193-213. 24. Peter K, Schwartz M, Ylanne J, et al. Induction of fibrinogen binding and platelet aggregation as a potential intrinsic property of various glycoprotein IIb/IIIa inhibitors. Blood 1998;92:3240-9.
Background Despite the increasingly prevalent role of platelet glycoprotein (GP) IIb/IIIa receptor inhibitors in acute coronary syndromes and percutaneous coronary interventions, the incidence and clinical relevance of thrombocytopenia occurring with their use remain unclear.
Methods We identified 8 placebo-controlled, randomized, large trials of GP IIb/IIIa receptor inhibitors reporting the incidence of thrombocytopenia, grouped by severity. The clinical courses of 42 patients with GP IIb/IIIa-related thrombocytopenia in these studies and other case reports were reviewed for bleeding complications.
Results Abciximab increased mild thrombocytopenia compared with placebo (4.2% vs 2.0%; P < .001; odds ratio 2.14) and increased severe thrombocytopenia compared with placebo (1.0% vs 0.4%; P = .01; odds ratio 2.48). Smallmolecule IIb/IIIa inhibitors did not significantly increase mild or severe thrombocytopenia compared with placebo. Mild thrombocytopenia occurred more frequently in acute coronary syndrome trials than in coronary intervention trials, even in patients not receiving any IIb/IIIa inhibitors. No major bleeding sequelae were reported in 23 patients with severe thrombocytopenia or in 19 patients with profound thrombocytopenia.
Conclusions Abciximab, but not eptifibatide or tirofiban, increases the incidence of thrombocytopenia compared with placebo in patients also treated with heparin. Thrombocytopenia associated with GP IIb/IIIa inhibition does not routinely lead to severe bleeding complications. (Am Heart J 2000;140:206-11.)
Platelets play a central role in initiating and mediating ischemia and related complications during acute coronary syndromes (ACS) and percutaneous coronary interventions (PCI). Platelet glycoprotein (GP) IIb/IIIa receptor inhibitors block the final common pathway to platelet aggregation and have played an increasing role in the management of ACS and PCI. However, these agents may cause internal bleeding and bleeding at the sites of vascular injury. Thrombocytopenia occurring with GP IIb/IIIa inhibitors has not been well-characterized. The relative propensity of various GP IIb/IIIa inhibitors to cause thrombocytopenia may affect their acceptance in clinical practice. We determined the incidence of thrombocytopenia occurring with different GP IIb/IIIa receptor inhibitors and its relation to clinically significant bleeding. From Geisinger Medical Center, Penn State Geisinger Health System, and The Janet Weis Research Center. Presented in part at the American College of Cardiology 48th Annual Scientific Sessions, March 9, 1999, New Orleans, La. Submitted December 27, 1999; accepted March 29, 2000. Reprint requests: James C. Blankenship, MD, Department of Cardiology 21-60, Geisinger Medical Center, Danville, PA 17822. E-mail: [email protected] Copyright © 2000 by Mosby, Inc. 0002-8703/2000/$12.00 + 0 4/1/107554 doi:10.1067/mhj.2000.107554
Methods Patients All major prospective, placebo-controlled, randomized, multicenter trials of intravenous GP IIb/IIIa receptor inhibitors were identified by a MEDLINE search. Only trials that clearly defined and documented levels of thrombocytopenia were included (Table I).1-8 Multicenter trials of lamifiban were excluded because they did not report specific levels or incidence of thrombocytopenia. Several trials1,4,5,8 excluded patients with baseline thrombocytopenia platelet count below 90,000 to 100,000/µL; others excluded patients with hemorrhagic diathesis but did not list specific exclusion criteria based on baseline platelet counts.2,3,6,7 Reports of large IIb/IIIa inhibitor trials were reviewed to identify descriptions of patients with severe thrombocytopenia, and a MEDLINE search was used to identify case reports or small studies of thrombocytopenia induced by GP IIb/IIIa inhibitors.
Definitions of thrombocytopenia Definitions of various levels of thrombocytopenia are not uniform. However, most studies define mild thrombocytopenia as a platelet count of <90,000 to 100,000/µL, severe thrombocytopenia <50,000/µL, and profound thrombocytopenia as <20,000/µL. The incidence of mild and severe thrombocytopenia in the selected studies was identified for each type of IIb/IIIa receptor inhibitor (abciximab, eptifibatide, and tirofiban) (Table I).
American Heart Journal Volume 140, Number 2
Statistical analysis The conditional test based on the odds ratio (OR) was used to compare the incidence of thrombocytopenia in patients treated with IIb/IIIa receptor inhibitor versus placebo.9 StatExact version 3.1 (Cytel Software, Cambridge, Mass) was used to calculate exact confidence intervals and P values. Additional analyses with these statistical tests aggregated trials by different characteristics (eg, ACS vs PCI, abciximab vs small-molecule drugs) to determine whether these differences in trial design affected the incidence of thrombocytopenia in patients receiving placebo/heparin and in patients receiving IIb/IIIa inhibitor/heparin. Statistical comparisons were not made across aggregated groups (eg, incidence of thrombocytopenia in heparin/placebo groups in ACS trials vs heparin/placebo groups in PCI trials). We avoided such statistical comparisons because patients admitted to one trial may have differed systematically from those enrolled in another regarding factors that might affect their risk of thrombocytopenia. Because of the multiple comparisons made among groups, conclusions about statistical significance were made only with extreme caution.
Results Eight studies that fit the inclusion criteria were identified (Table I). Table II shows the incidence of mild thrombocytopenia in patients receiving specific types of IIb/IIIa receptor inhibitor with heparin versus placebo with heparin. Abciximab increased the incidence of mild thrombocytopenia compared with the placebo group (4.2% vs 2.0%; P < .001; OR 2.13). Eptifibatide or tirofiban with heparin did not increase mild thrombocytopenia compared with placebo with heparin (OR 0.99). Table III shows the incidence of severe thrombocytopenia in the same groups as in Table II. Patients receiving abciximab with heparin had more than twice the incidence of severe thrombocytopenia than those receiving placebo with heparin (1.0% vs 0.4%; P = .01; OR 2.48). Eptifibatide or tirofiban with heparin did not cause a significant excess of severe thrombocytopenia compared with placebo with heparin (0.3% vs 0.2%, P = .16). In the above analyses, abciximab data came predominantly from PCI trials, and small-molecule drug data came predominantly from ACS trials. To determine if the lack of thrombocytopenia with small-molecule inhibitors was obscured by the design of ACS trials, we compared small-molecule inhibitor ACS trials with small-molecule inhibitor PCI trials (Table IV). Smallmolecule IIb/IIIa inhibitors did not cause a significant excess of thrombocytopenia in either coronary intervention trials or ACS trials. (This analysis could not be performed with abciximab because no data were available from abciximab ACS trials.) To exclude type of trial (ACS vs PCI) as a possible confounding factor, these analyses were repeated excluding all ACS trials (Table V). PCI trials were aggregated by type of IIb/IIIa inhibitor (abciximab vs smallmolecule inhibitors). As before, abciximab, but not
Dasgupta et al 207
the small-molecule inhibitors, was associated with an excess of mild and severe thrombocytopenia compared with placebo. Table II shows that patients receiving placebo (not receiving any IIb/IIIa inhibitor) developed mild thrombocytopenia in small-molecule IIb/IIIa trials (4.6%) more often than in abciximab trials (2.0%). This variation in the incidence of thrombocytopenia among placebo groups from different trials raised the possibility that trial design affected the incidence of thrombocytopenia. To detect such an effect, trials were aggregated by type of trial (ACS vs PCI) rather than by type of IIb/IIIa inhibitor (Table VI). Patients receiving placebo had mild thrombocytopenia develop in 5.8% of ACS trials versus 1.9% in PCI trials. This analysis was repeated excluding abciximab trials; when small-molecule IIb/IIIa inhibitor ACS trials were compared with small-molecule IIb/IIIa inhibitor PCI trials (Table IV), placebo groups in ACS trials had 3-fold more mild thrombocytopenia than placebo groups in PCI trials (5.8% vs 1.7%). Table V provides additional data indicating that differences in incidence of thrombocytopenia in placebo groups among studies were caused by the type of trial (ACS vs PCI), not type of IIb/IIIa inhibitor used. When ACS trials were excluded, abciximab PCI trials and small-molecule IIb/IIIa PCI trials produced very similar incidences of mild thrombocytopenia in patients receiving placebo with heparin (2.0% vs 1.7%). The outcomes of 42 patients, 23 with severe thrombocytopenia and 19 with profound thrombocytopenia, are listed in Table VII.4,5,10-18 None of these involved retreatment with abciximab. No major bleeding complication was reported in any of the patients despite levels of thrombocytopenia as low as <10,000/µL. Thrombocytopenia was transient and resolved either spontaneously or with platelet transfusions.
Discussion The most important finding of our analysis is the significant increase in thrombocytopenia (mild and severe) when used with abciximab and heparin compared with placebo and heparin. Although there was a trend toward a similar increase in severe thrombocytopenia with other GP IIb/IIIa receptor inhibitors, it was not statistically significant. Patients treated with placebo had higher incidence of mild thrombocytopenia in ACS studies (compared with PCI trials), suggesting that the design of these studies may have increased the incidence of thrombocytopenia regardless of the IIb/IIIa inhibitor used in the trial. This trend was not observed for severe thrombocytopenia. It is likely that the longer infusions of heparin used in ACS trials account for more frequent mild thrombocytopenia. Heparin-induced thrombocytopenia rarely causes profound thrombocytopenia. Consistent with that, there was no increase in profound thrombocytopenia
American Heart Journal August 2000
208 Dasgupta et al
Table I. Studies reporting thrombocytopenia with GP IIb/IIIa receptor inhibitors
Study CAPTURE1 EPIC2 EPILOG3 PRISM4 PRISM-PLUS5 PURSUIT6 IMPACT II7 RESTORE8
Study drug
Patient population
Placebo + heparin (n)
GP IIb/IIIa + heparin (n)
A A A T T E E T
PCI PCI PCI ACS ACS ACS PCI PCI
635 696 939 1616 797 4696 1285 1070
630 1403 1853 — 773 4679 2586 1071
GP IIb/IIIa (n) — — — 1616 345 — — —
Empty cells reflect drug combinations not used in the trials. A, Abciximab, T, tirofiban, E, eptifibatide.
Table II. Mild thrombocytopenia according to drug groups Treatment group Heparin + any GP IIb/IIIa Heparin + abciximab Heparin + small-molecule GP IIb/IIIa GP IIb/IIIa alone‡
Incidence in GP IIb/IIIa + heparin groups
n* 7 3 4 1
4.5% 4.2% 4.6% 1.1%†
Incidence in placebo + heparin groups
OR
95% confidence interval for OR
Exact P value
4.1% 2.0% 4.6% 0.4%†
1.11 2.13 0.99 3.02
(0.96-1.25) (1.52-3.04) (0.86-1.15) (1.15-9.32)
.17 <.001 .94 .02
Incidence in placebo + heparin groups
OR
95% confidence interval for OR
Exact P value
0.2% 0.4% 0.2% 0.1%
2.31 2.48 1.73 3.01
*Number of studies contributing to comparison. †These patient groups received GP IIb/IIIa inhibitors without heparin. ‡PRISM (using tirofiban in patients with ACS).
Table III. Severe thrombocytopenia according to drug groups Treatment group Heparin + any GP IIb/IIIa Heparin + abciximab Heparin + small-molecule GP IIb/IIIa GP IIb/IIIa alone†
n* 7 3 4 1
Incidence in GP IIb/IIIa + heparin groups 0.5% 1.0% 0.3% 0.4%*
(1.38-3.98) (1.18-5.85) (0.83-3.79) (0.54-30.50)
<.001 .01 .16 .29
These patient groups received GP IIb/IIIa inhibitors without heparin. *Number of studies contributing to comparison. †PRISM (using tirofiban in patients with ACS).
in ACS versus PCI trials. An increase in heparin-induced thrombocytopenia from the longer infusions used in ACS trials (versus PCI trials) might explain in part the higher incidence of thrombocytopenia in patients receiving IIb/IIIa inhibitors in ACS trials (vs PCI trials). Our study documents the relatively benign clinical course of thrombocytopenia associated with first exposure to GP IIb/IIIa receptor inhibitors. Access site hematomas and minor oozing were the most frequent
bleeding complications. One patient had hematemesis without other complications.15 Thrombocytopenia was transient and either improved spontaneously or with platelet transfusions (in one instance with additional intravenous immunoglobulin and steroids).10 The cause for the observed difference in incidence of thrombocytopenia induced by abciximab as opposed to eptifibatide or tirofiban is unclear. Abciximab induces an antibody response (human antichimeric antibod-
American Heart Journal Volume 140, Number 2
Dasgupta et al 209
Platelet count <90,000-100,000/µL Placebo + heparin (n)
GP IIb/IIIa + heparin (n)
8 24 14 6 6 315 31 10
<50,000/µL GP IIb/IIIa (n)
35 82 47 — 15 318 71 13
Placebo + heparin (n)
GP IIb/IIIa + heparin (n)
GP IIb/IIIa (n)
0 5 4 2 2 4 5 1
12 14 12 — 4 9 9 2
— — — 6 — — — —
— — — 18 — — — —
Table IV. Thrombocytopenia in small-molecule IIb/IIIa trials aggregated by ACS versus PCI trials Platelet count (per µL)
Type of trial
No. of studies
No. of patients
<90,000-100,000 <90,000-100,000 <50,000 <50,000
ACS PCI ACS PCI
2 2 2 2
10,945 6012 10,945 6012
Incidence in GP IIb/IIIa + heparin groups 6.1% 2.3% 0.2% 0.3%
Incidence in placebo + heparin groups 5.8% 1.7% 0.1% 0.3%
OR
95% Confidence interval for OR
Exact P value
1.05 1.33 2.19 1.18
(0.89-1.23) (0.90-1.99) (0.77-7.02) (0.40-3.90)
.59 .16 .16 .95
Exact P value .16 <.001 .95 .01
Table V. Thrombocytopenia in PCI trials by type of IIb/IIIa inhibitor Platelet count (per µL)
Type of drug
No. of studies
No. of patients
Incidence in GP IIb/IIIa + heparin groups
Incidence in placebo + heparin groups
OR
95% Confidence interval for OR
<90,000-100,000 <90,000-100,000 <50,000 <50,000
Small-molecule Abciximab Small-molecule Abciximab
2 3 2 3
6012 6156 6012 6156
2.3% 4.2% 0.3% 1.0%
1.7% 2.0% 0.3% 0.4%
1.33 2.13 1.18 2.48
(0.90-1.99) (1.52-3.04) (0.40-3.90) (1.18-5.85)
ies).19,20 These antibodies may lead to increased platelet consumption and contribute to higher incidence of thrombocytopenia with abciximab. Other mechanisms of thrombocytopenia with abciximab may include the development and subsequent removal of aggregates of platelets and c7E3 Fab, binding of preexisting immunoglobulin G antibodies to abciximab-induced new platelet epitopes (ligand-induced binding sites), and other kinds of platelet modifications resulting from interaction with abciximab.21-24 The contribution, if any, of these mechanisms to the observed difference
in thrombocytopenia caused by abciximab versus other GP IIb/IIIa receptor inhibitors is currently unknown. Limitations of this study include those of any metaanalysis. It is retrospective, comparing trials that differ to some extent in drugs, treatment protocols, monitoring protocols, and definitions. Several large trials of GP IIb/IIIa receptor inhibitors could not be included because they did not report the incidence of thrombocytopenia. Because of the multiple comparisons made among groups, conclusions about statistical significance should be drawn very cautiously. Ideally, the rel-
American Heart Journal August 2000
210 Dasgupta et al
Table VI. Thrombocytopenia: ACS versus PCI trials Platelet count (per µL)
Type of trial
No. of studies
No. of patients
Incidence in GP IIb/IIIa + heparin groups
Incidence in placebo + heparin groups
<90,000-100,000 <90,000-100,000 <50,000 <50,000
ACS PCI ACS PCI
2 5 2 5
10,945 12,168 10,945 12,168
6.1% 3.3% 0.2% 0.6%
5.8% 1.9% 0.1% 0.3%
OR
95% Confidence interval for OR
Exact P value
1.05 1.77 2.19 2.01
(0.89-1.23) (1.38-2.30) (0.77-7.02) (1.11-3.86)
.59 <.001 .16 .02
Table VII. Clinical consequences of GP IIb/IIIa-induced severe and profound thrombocytopenia Study PRISM4 PRISM-PLUS5 EPIC10 CAPTURE1 Kereiakes et al11
Patients with platelet count <50,000/µL 6 4
2 10
Berkowitz et al12 Mendes et al13 Jenkins et al14 Jubelirer et al15 Ferrari et al16 Joseph et al17 Elmi et al18
Patients with platelet count <20,000/µL
2 4 4 1 1 3
3 1 1
ative contribution of IIb/IIIa inhibitors and heparin to thrombocytopenia could be assessed by multiple logistic regression analysis of the original trial databases pooled together, but without access to these databases this approach is impossible. The individual reports of GP IIb/IIIa-associated thrombocytopenia do not allow comparison of bleeding sequelae in patients who did not receive these drugs. However incomplete, these cases suggest that severe bleeding is unusual (none of 42 patients) with GP IIb/IIIa-associated thrombocytopenia, and it highlights the need for definition and reporting of thrombocytopenia in large GP IIb/IIIa trials. In summary, mild and severe thrombocytopenia occur more frequently in patients treated with abciximab and heparin than with placebo and heparin. Tirofiban and eptifibatide alone or in combination with heparin did not result in a statistically significant higher incidence of thrombocytopenia than heparin use alone. ACS trials tended to report higher incidences of mild thrombocytopenia (compared with coronary intervention trials), perhaps because of longer heparin infusions producing heparin-induced thrombocytopenia. Severe and profound thrombocytopenia from GP IIb/IIIa receptor inhibitors
Bleeding complications and interventions No clinical sequelae No sequelae No bleeding; platelet transfusion in both, intravenous immunoglobulin + steroid in one No bleeding; 5 received platelet transfusions 3 Groin, 1 forearm hematoma, no major bleeding; all received platelet transfusions, 2 received 2 U red blood cell transfusion each No major bleeding; small oozing + groin hematoma in 1 No bleeding; received platelet transfusion No bleeding Hematoma of groin, forearm, and scrotum; hematemesis in 1 (platelet and blood transfusion [2 U] in this patient only) No bleeding No bleeding No bleeding, prophylactic platelet transfusion given
does not often cause severe bleeding. Future trials should clearly define and report mild, severe, and profound thrombocytopenia (platelet counts <100,000/µL, <50,000/µL, and <20,000/µL, respectively) to facilitate comparison of GP IIb/IIIa receptor inhibitors. Future trials should also evaluate patients with thrombocytopenia for heparin-induced thrombocytopenia.
References 1. The CAPTURE investigators. Randomized placebo-controlled trial of abciximab before and during coronary intervention in refractory unstable angina: the CAPTURE study. Lancet 1997;349:1429-35. 2. The EPIC investigators. Use of a monoclonal antibody directed against the platelet glycoprotein IIb/IIIa receptor in high risk coronary angioplasty. N Engl J Med 1994;330:956-61. 3. The EPILOG investigators. Platelet glycoprotein IIb/IIIa receptor blockade and low dose heparin duration percutaneous coronary revascularization. N Engl J Med 1997;336:1689-96. 4. The PRISM study investigators. A comparison of aspirin plus tirofiban with aspirin plus heparin for unstable angina. N Engl J Med 1998; 338:1498-505. 5. The PRISM-PLUS study investigators. Inhibition of the platelet glycoprotein IIb/IIIa receptor with tirofiban in unstable angina and nonQ-wave myocardial infarction. N Engl J Med 1998;338:1488-97.
American Heart Journal Volume 140, Number 2
6. The PURSUIT trial investigators. Inhibition of platelet glycoprotein IIb/IIIa with eptifibatide in patients with acute coronary syndromes. N Engl J Med 1998;339:436-43. 7. The IMPACT-II investigators. Randomized placebo- controlled trial of eptifibatide on complications on percutaneous coronary interventions: IMPACT-II. Lancet 1997;347:1422-8. 8. The RESTORE investigators. Effect of platelet glycoprotein IIb/IIIa blockade with tirofiban on adverse cardiac events in patients with unstable angina or acute myocardial infarction undergoing coronary angioplasty. Circulation 1997;96:1445-53. 9. Gart J. Point and interval estimation of the common odds ratio in the combination of 2X2 tables with fixed margins. Biometrika 1970; 57:471-5. 10. The EPIC study group. Occurrence and clinical significance of thrombocytopenia in a population under going high-risk percutaneous revascularization. J Am Coll Cardiol 1998;32:311-9. 11. Kereiakes DJ, Essell JH, Abbottsmith CW, et al. Abciximab-associated profound thrombocytopenia: therapy with immunoglobulin and platelet transfusion. Am J Cardiol 1996;78:1161-3. 12. Berkowitz SD, Harrington RA, Rund MM, et al. Acute profound thrombocytopenia after c7E3 Fab (abciximab) therapy. Circulation 1997;95:809-13. 13. Mendes V, Resende M, Santos JM, et al. Trombocitopenia grave associada a terapeutica com abciximab-a proposito de um caso clinico. Revista Portuguesa de Cardiologia 1998;17:609-13. 14. Jenkins LA, Lau S, Crawford M, et al. Delayed profound thrombocytopenia after c7E3 Fab (abciximab) therapy [letter]. Circulation 1998;97:1214-5. 15. Jubelirer SJ, Koenig BA, Bates MC. Acute profound thrombocytopenia following c7E3 Fab (abciximab) therapy: case reports, review of literature and implications for therapy. Am J Hematol 1999;61:205-8.
Dasgupta et al 211
16. Ferrari E, Thiry M, Touati C, et al. Acute profound thrombocytopenia after c7E3 Fab therapy [letter]. Circulation 1997;95: 3809-10. 17. Joseph T, Marco J, Grogorini L. Acute profound thrombocytopenia after abciximab therapy during coronary angioplasty. Clin Cardiol 1998;21:851-2. 18. Elmi F, Oza R, Mascarenhas DAN. Acute profound thrombocytopenia without bleeding complications after abciximab administration. J Invas Cardiol 1999;11:313-5. 19. Berkowitz SD, Sane DC, Sigmon KN, et al. Occurrence and clinical significance of thrombocytopenia in a population undergoing highrisk percutaneous coronary revascularization. J Am Coll Cardiol 1998;32:311-9. 20. Jordan RE, Wagner CL, Mascelli MA, et al. Preclinical development of c7E3 Fab: a mouse/human chimeric monoclonal antibody fragment that inhibits platelet function by blockade of GP IIb/IIIa receptors with observations on the immunogenicity of c7E3 Fab in humans. In: Horton MA, editor. Adhesion receptors as therapeutic targets. New York: CRC Press; 1996. p 281-302. 21. Frelinger AL III, Cohen I, Plow EF, et al. Selective inhibition of integrin function by antibodies specific for ligand-occupied receptor conformers. J Biol Chem 1990;265:6346-52. 22. Bednar B, Bednar RA, Cook JJ, et al. Drug-dependent antibodies against GPIIb/IIIa induce thrombocytopenia [abstract]. Circulation 1996;94 suppl I:I-99. 23. Coller BS, Scudder LE, Beer J, et al. Monoclonal antibodies to platelet glycoprotein IIb/IIIa as antithrombotic agents. Ann NY Acad Sci 1991;614:193-213. 24. Peter K, Schwartz M, Ylanne J, et al. Induction of fibrinogen binding and platelet aggregation as a potential intrinsic property of various glycoprotein IIb/IIIa inhibitors. Blood 1998;92:3240-9.