Immune thrombocytopenia (ITP) is an autoimmune condition characterized by isolated severe thrombocytopenia. Hemorrhagic tendency is the main manifestation of ITP, but the risk of thrombosis in patients with ITP is thought to be 5%, predominantly causing venous thromboembolism. Reference Thachil, Callaghan and Martlew1 Risk of thrombosis is similar to that associated with malignancy. There is lack of evidence on how to manage acute ischemic stroke and manage secondary prevention in patients with ITP. This is of particular importance as the standard of care for acute ischemic stroke includes a prompt initiation of reperfusion therapies and subsequent use of antiplatelets or anticoagulant agents, which can increase the risk of major bleeding. Concern for potential life-threatening hemorrhage in ITP patients influences decision-making in acute stroke management.
We present two patients with ITP and acute ischemic stroke who received tissue plasminogen activator (tPA) and endovascular thrombectomy (EVT). In addition, we performed a review of the literature to summarize the available evidence.
Case Series
Case 1
A 61-year- old female was brought to hospital after a witnessed sudden onset of right-sided weakness and global aphasia that started 90 min prior. She had been diagnosed with ITP more than a decade prior to her stroke onset. She had been treated with corticosteroid and intravenous rituximab, and ultimately underwent splenectomy 13 years ago. Despite this, she failed to achieve complete long-term remission. At the time of her stroke she was not on antiplatelet or ITP-specific therapy. Two months prior to stroke onset her platelet count was 28 × 109/l. Her past medical history included coronary artery disease, dyslipidemia, hypertension, breast cancer in remission for several years, and a remote pulmonary embolism which was managed with an inferior vena cava (IVC) filter and no anticoagulation. Upon arrival to the emergency room, she was in atrial fibrillation (AF) but hemodynamically stable. Her blood pressure was 165/80 mmHg and heart rate was 85 beats per minute (bpm). Blood glucose level was 6.5 mmol/l (117 mg/dl). Neurological examination revealed drowsiness, global aphasia, and right-sided dense hemiparesis. Her National Institute of Health Stroke Scale (NIHSS) was 16. Computed tomography (CT) of the head demonstrated left middle cerebral artery (MCA) hyperdense sign and early acute ischemic changes within the left basal ganglia and inferior left insula (Figure 1(A) and (B)). CT angiogram (CTA) head and neck showed thrombotic occlusion of the left carotid terminus extending into the left M1 segment of MCA (Figure 1(C) and (D)) with good collaterals and no significant atherosclerotic changes. Her platelets at presentation were 48 × 109/l. Given the significant neurological deficits, timely presentation to emergency department, and her high prehospitalization functional capacity, the decision was made with the patient’s family to proceed with thrombolysis using tPA. This was followed by a single-pass successful EVT using combined approach with stent retriever and aspiration assisted by balloon guide catheter (Figure 1(E) and (F)). The door to successful recanalization time of the thrombolysis in cerebral infarction (TICI) 2B was approximately 120 min. Follow-up CT head after 24 h showed the expected evolution of ischemic stroke, but there was no hemorrhagic complication. Platelet transfusion was not considered prior to thrombolysis/EVT or required during hospital admission. Four days after thrombolysis and EVT, her NIHSS was 0 and her modified Rankin Score (mRS) was 1 for very mild short-term memory difficulties. While in hospital, she was assessed by hematology and acetylsalicylic acid (ASA) for secondary stroke prevention was recommended, but no specific immunotherapy for ITP was suggested. Further assessment and discussion of risk and benefit of anticoagulation for AF was arranged with outpatient hematology after discharge. At 6 months of follow-up, there was no hemorrhagic or thrombotic complication, and her mRS remained 1. Unfortunately, the patient did not return for hematology follow-up on a number of occasions to assess the risks and benefits of switching ASA to anticoagulation. She was therefore continued on ASA for secondary stroke prevention.
Figure 1: Diagnostic imaging and therapeutic intervention for the first case reported. (A) Axial CT head shows left MCA hyperdense sign indicating a fresh clot. (B) Axial CT head reveals early ischemic changes in the left insula and basal ganglia as indicated by the arrows. (C and D) Axial and coronal sections of CTA show thrombotic occlusion of the left carotid terminus extending into the left M1 segment of MCA. (E) Digital subtraction angiogram further demonstrates the same finding shown in CTA. (F) Digital subtraction angiogram shows revascularization after a successful endovascular thrombectomy. (G) Diffusion weighted imaging of the brain 24 h after the ischemic stroke reveals a diffusion restriction in the left insula and basal ganglia.
Case 2
A 75-year-old male was brought to hospital after he was found to have left-sided weakness and slurred speech that started an hour prior. He had been diagnosed with ITP a few years earlier with a remote history of life-threatening gastrointestinal hemorrhage and epistaxis a year ago. The cause of the bleeding was thought to be secondary to critical thrombocytopenia as gastro-endoscopic evaluations were negative for malignancy and ulcer. Four months prior to this presentation, he experienced epistaxis with platelet count of 75 × 109/l. He was treated with corticosteroids and IVIG for ITP 6 months prior, and 2 months prior to presentation underwent splenectomy. His last platelet count was 349 × 109/l 2 days before stroke onset. His past medical history also included coronary artery disease, hypertension, and dyslipidemia. At the time of his stroke he was on ASA but not on any ITP-specific therapy. His blood pressure was 175/95 mmHg and heart rate was 90 bpm. His blood glucose level was 7 mmol/l (126 mg/dl). Neurological examination revealed left-sided neglect, right gaze forced deviation, left-sided weakness, decreased sensation, and dysarthria. NIHSS was 14. His CT of the head revealed a right distal MCA hyperdense sign in the sylvian fissure correlating with right M2 occlusion (Figure 2(A)). CTA of the head and neck demonstrated proximal occlusion of M2 segment of the right MCA (Figure 2(B) and (C)) with good collaterals and no significant atherosclerotic changes. Given the significant neurological deficits, the decision was made to proceed with thrombolysis using tPA. He then underwent a single pass successful EVT using combined approach with stent retriever and aspiration assisted by balloon guide catheter (Figure 2(D) and (E)). The door to successful recanalization time of TICI 3 was less than 60 min. He tolerated the procedure well with no immediate complications. Shortly after the procedure, his platelet count result became available, and it was 366 × 109/l, and his NIHSS improved to 9. His blood pressure remained stable ranging from 130/85 to 150/95 before and after intervention. He did not require acute antihypertensive therapy. After 2 h of thrombolysis/EVT, he developed hematemesis, drowsiness, and increasing left-sided weakness. Repeated CT head revealed a right frontoparietal, moderately large, intracerebral hemorrhage (ICH) with some subarachnoid hemorrhage (SAH) (Figure 2(F), (G), and (H)). His coagulation parameters were within normal range. INR was 1.1 units, PTT was 36 s, and fibrinogen was 3.2 g/l. In consultation with transfusion medicine, he was given 16 units of cryoprecipitate and one pooled unit of platelets. He was not started on corticosteroid or specific immunotherapy for ITP as it was felt that he was not in ITP relapse. He suffered from additional medical complications, including aspiration pneumonia, congestive heart failure, and AF. Upon discussion with family, goals of care were changed to palliative care and the patient ultimately died.
Figure 2: Diagnostic imaging, therapeutic intervention, and follow-up scan for the second case reported. (A) Axial CT head reveals a right distal MCA hyperdense sign in the sylvian fissure correlating with right M2 occlusion. (B and C) Axial and coronal CTA show proximal occlusion of M2 segment of the MCA. (D) Digital subtraction angiogram further demonstrates the same finding shown in CTA. (E) Digital subtraction angiogram shows revascularization after a successful EVT. (F, G, and H) Axial CT head after tPA EVT reveals right frontoparietal SAH and large ICH surrounded by edema.
Discussion
Summary of Reported Cases
The incidence of ITP in the general population is estimated to be 2 to 5 per 100,000 people. Reference Neunert, Terrell, Arnold, Buchanan, Cines and Cooper2 A literature review reveals 27 case reports relating to ischemic stroke in patients with ITP (Table 1). Most of those patients had at least one vascular risk factor in addition to their ITP disorder. The mode of onset in all cases varied between acute and subacute presentations. None of the patients received thrombolysis or an endovascular procedure except a single report of intraarterial urokinase given to a patient with ischemic stroke with platelet count of 84 × 109/l. Reference Rhee, Choi, Kim and Shin3 More than half of the patients were started on antiplatelet for acute management and/or secondary prevention. Hemorrhagic transformation (HT) occurred in four patients, three of them were on antiplatelets and one was on anticoagulation. Two of the patients who developed HT had significant thrombocytopenia and the other two had a platelet count >100 × 109/l.
Table 1: Clinical, imaging, and therapeutic characteristics in 27 patients reported in the literature with ITP and ischemic stroke
IA = intraarterial; ICA = internal carotid artery; IV = intravenous; MCA = middle cerebral artery; MRA = magnetic resonance arteriography; MRI = magnetic resonance imaging; NR = not reported; PCA = posterior cerebral artery; SCA = superior cerebellar artery.
Mechanism of Thrombosis
The general pathophysiology of ITP is that of an acquired autoimmune disorder, which can be primary or secondary, causing platelet destruction and impaired production. Reference Neunert, Terrell, Arnold, Buchanan, Cines and Cooper2 Multiple hypotheses have been proposed to explain the mechanism of thrombosis in patients with ITP. The current leading hypothesis revolves around platelet microparticles (PMPs). PMPs are secreted by both activated and destroyed platelets as they are the natural elements that promote thrombosis for hemostatic control, but they also may, in part, be responsible for pathological thrombosis in patients with ITP. Compared to healthy controls, elevated levels of PMPs are detected in patients with ITP and concurrent ischemic events. Reference Jy, Horstman, Arce and Ahn4 Elevated levels of PMPs have also been found in patients with ITP and vascular dementia due to ischemic small vessel disease – these patients also had higher platelet counts as well as more often had splenectomy. Reference Ahn, Horstman, Jy, Jimenez and Bowen5 PMPs are not alone to blame for thrombosis in ITP; it is thought that during immune-related platelet destruction there is a large proportion of immature activated platelets released from bone marrow, as well as large platelet–leukocyte–monocyte aggregates circulating and endothelium activating antibodies which all contribute to an increased risk of thrombosis. Reference Thachil, Callaghan and Martlew1,Reference Rhee, Choi, Kim and Shin3,Reference Ahn, Horstman, Jy, Jimenez and Bowen5,Reference Aledort, Hayward, Chen, Nichol, Bussel and Group6 Physiological nitric oxide (NO) in vessel endothelium in healthy individuals prevents platelet adhesion to vessel walls, and it is possible that in ITP, NO is depleted as a consequence of immune activation and also contributes to the prothrombotic state in ITP. Reference Thachil, Callaghan and Martlew1,Reference Freedman and Loscalzo7 Neither of the patients presented in our case series were actively receiving intravenous immunoglobulin (IVIG) to treat their ITP, but IVIG is a known prothrombotic medication. Both of our patients had undergone splenectomy, which has consistently been shown to cause 2 times the risk of venous thromboembolism than the general population, and an insignificant increased risk of arterial thrombosis. Reference Rodeghiero8 In our cases, we cannot be certain about the exact mechanism of ischemic stroke as both of them had AF and conventional vascular risk factors in addition to their ITP condition.
Clinical Decision-Making
We faced a management dilemma for our two patients given the insufficient data that represent the outcome of patients with ITP and ischemic stroke. Patients with significant thrombocytopenia (<100 × 109/l) are already at increased risk of major bleeding and have been uniformly excluded from all major acute ischemic stroke clinical trials. The paucity of data surrounding the outcomes of these patients makes it difficult to weigh the risks and benefits of potentially lifesaving and disability-preventing therapies. Specific data related to the risk of bleeding after thrombolysis in patients with ischemic stroke and ITP are lacking. In general, the 5-year cumulative rate of spontaneous ICH in adult ITP patients is 1.4% which is 3 times higher than patients without ITP; however, all patients with reported spontaneous ICH had platelets of equal to or greater than 30 × 109/l. Reference Norgaard, Jensen, Engebjerg, Farkas, Thomsen and Cha9,Reference Neunert, Noroozi, Norman, Buchanan, Goy and Nazi10 There are other factors than ITP that increase the risk of intracerebral bleeding postreperfusion. AF, elevated blood pressure, taking an antiplatelet agent prior to EVT and thrombolysis, and statin use have all been associated with higher risks of symptomatic HT. Reference van Kranendonk, Treurniet, Boers, Berkhemer, van den Berg and Chalos11 We cannot establish that the cause of postreperfusion ICH in our second patient was due to ITP. He had multiple additional factors other than his ITP including the fact he was on an antiplatelet agent at presentation, had a new diagnosis of AF, and a history of past severe bleeding events. ITP may have contributed to his risk for bleeding despite having normal platelet count.
Acute Stroke Therapy in Patient with ITP
The use of thrombolysis in patients with ITP
The American Heart Association/American Stroke Association (AHA/ASA) guideline recommended that tPA for patients with acute stroke and a clinical history of potential bleeding diathesis may be considered on a case-by-case basis because the safety and efficacy of tPA in this situation is unknown. Reference Powers, Rabinstein, Ackerson, Adeoye, Bambakidis and Becker12,Reference Demaerschalk, Kleindorfer, Adeoye, Demchuk, Fugate and Grotta13 Significant thrombocytopenia regardless of the cause is considered a relative contraindication for tPA for acute ischemic stroke per AHA/ASA and the Canadian Stroke Best Practice Recommendations guidelines. Reference Powers, Rabinstein, Ackerson, Adeoye, Bambakidis and Becker12,Reference Boulanger, Lindsay, Gubitz, Smith, Stotts and Foley14 The relationship between platelet count and hemorrhage risk is unknown. There have been no randomized trials or prospective studies to evaluate the risk of hemorrhage in patients with acute ischemic stroke and significant thrombocytopenia. The threshold of <100 × 109/l for platelets being a contraindication for thrombolysis was derived from an expert panel consensus. In a study that looked at patients in the Thrombolysis in Stroke Patients (TRISP) registry, among 7,533 tPA-treated stroke patients, 44 had significant thrombocytopenia (< 100 × 109/l). Reference Gensicke, Al Sultan, Strbian, Hametner, Zinkstok and Moulin15 Three patients (6.8%) developed symptomatic ICH (sICH) using the Second European Cooperative Acute Stroke Study (ECASS II) criteria. Reference Hacke, Kaste, Fieschi, von Kummer, Davalos and Meier16 In the same study, the overall risks of sICH, poor functional outcome, and mortality did not differ significantly from those patients with a platelet count >100 × 109/l. It was found that every decrease in platelet count by 10,000/l increased the risk of sICH by 2–5%, yet this was not associated with poor outcome or mortality. In addition, there are 27 patients in multiple studies and case reports who received tPA with significant thrombocytopenia. Reference Frank, Grotta, Alexandrov, Bluhmki, Lyden and Meretoja17–Reference Bragin and Chen22 The detailed information about the cause of thrombocytopenia was not available in most cases. Two of these patients (7.7%) developed sICH. By compiling patients from all the studies, the risk sICH in significant thrombocytopenia post-tPA is 7%. Although the sample size is very small, the rate is not significantly different than sICH in patients with normal platelet counts treated with tPA. In in vitro studies, tPA has shown to inhibit platelet aggregation, but does not affect platelet activation, and perhaps the reports of bleeding post-tPA in individuals with ITP are not as high as expected. Reference Lu, Hu, Wei, Luo, Qiao and Geng23 This further challenges the justification of withholding thrombolysis for platelet counts <100 × 109/l. However, in ITP platelets may be dysfunctional for other reasons as discussed previously.
Performing EVT for patients with ITP
Similar to tPA, a specific data related to the safety of EVT in patients with ischemic stroke and ITP is lacking. Although EVT without thrombolysis in patients with ITP who present with the acute ischemic stroke and large vessel occlusion may be a safer option, the experience with EVT alone in these patients is also limited. Lower thresholds for thrombocytopenia regardless of the cause were used in the EVT studies, such as (Multi – MERCIİ: <30 × 109/l, MR-CLEAN: < 40 × 109/l, SWIFT PRIME and EXTEND IA:<100 × 109/l, DAWN and DEFUSE-3: <50 × 109/l). Reference Smith, Sung, Saver, Budzik, Duckwiler and Liebeskind24–Reference Albers, Marks, Kemp, Christensen, Tsai and Ortega-Gutierrez29 In the MERCI/Multi MERCI cohort, EVT was used in six patients with significant thrombocytopenia. Reference Nogueira and Smith30 Symptomatic ICH was noted in one patient (platelet count was 16 × 109/l), mild SAH in another patient (platelet count was 64 × 109/l), and four patients did not develop any hemorrhagic complications (platelets counts were 37 − 94 × 109/l). A recently published retrospective observational analysis from a single center has assessed the EVT in patients with thrombocytopenia. Reference Desai, Mehta, Morrison, Gross, Jankowitz and Jovin31 Fifteen patients with significant thrombocytopenia underwent EVT. The information for the cause of thrombocytopenia in each individual was not available. Symptomatic ICH occurred in one patient with no groin bleeding complication in any patients.
There are many factors that need to be considered when deciding to proceed with thrombolysis and EVT in patients with acute stroke. This includes premorbid functional status, presence of other medical comorbidities, severity of stroke, goals of care, and timely access to different therapeutic options. If patients are known to have ITP, consideration to the stability of the disease, last known platelet count and time of last seen well should be taken into account prior to initiating acute stroke therapy. It is reasonable to wait prior to initiating any therapy for the platelet count to return in patients who have active disease, have clinical stigmata of thrombocytopenia, or their most recent platelet count prior to presentation was <100 × 109/l. In patients who present within the thrombolysis window but who are felt not to be eligible for thrombolysis, or who are outside the window, they should be considered for EVT. If a patient is known to have ITP or who has recent unexplained low platelet count, urgent consultation with hematology is recommended prior to proceeding with any therapies, and decisions should be made in collaboration with stroke clinicians, hematologists, the patients and their families where possible.
The use of antiplatelet and anticoagulation agents in patients with ITP
Data on the use of antiplatelet and anticoagulation in ITP for ischemic stroke are lacking; thus, the strategy for stroke prevention should be individualized according to the stroke mechanism, comorbidities, and risk of hemorrhagic complications. No direct evidence exists to recommend a platelet threshold for safe antiplatelet or anticoagulant use, but a survey of International ITP experts reported a threshold of platelets >50 × 109/l for the use of both antiplatelets and anticoagulants in acute and long-term situations. Reference Pishko, Misgav, Cuker, Cines, George and Vesely32 This value is also accepted in the cardiology literature as they suggest using dual antiplatelet therapy with clopidogrel and aspirin after stenting provided platelets are at least >50 × 109/l. Reference McCarthy, Steg and Bhatt33 If antithrombotic agents are indicated for acute or long term, a platelet count of >50 × 109/l is recommended, and to hold them if platelet counts drop below this. Reference McCarthy, Steg and Bhatt33–Reference Mahawish, Pocock, Mangarai and Sharma35 In vitro studies have shown that both aspirin and P2Y12 inhibitors are able to block the prothrombotic effects of PMP, and thus either could be used in primary and secondary prevention. The stroke guidelines currently suggest aspirin as first line for secondary prevention. Reference Giacomazzi, Degan, Calabria, Meneguzzi and Minuz36 The decision for long-term antiplatelet and anticoagulation treatment should be individualized and take into account the presence or absence of other stroke risk factors.
A therapeutic algorithm for patients with ITP and thrombotic complications has been proposed. Reference Matzdorff and Beer34 Anticoagulation is not recommended if there is a life-threatening hemorrhage or hemorrhage that requires transfusion (WHO grade III/IV). Corticosteroid and IVIG may be administered to increase the platelet count to a safe level ≥50 × 109/l before starting anticoagulation or antiplatelet agents. Case reports of patients with ITP and acute stroke have demonstrated that if a stroke is deemed to be caused by ITP, and a decision for acute stroke therapy has been proposed, immunosuppressing agents such as IVIg and corticosteroids can safely be initiated in the acute phase of stroke, the thrombopoietin receptor agonists (TPO-RAs) can also be considered in place of immunosuppressants to increase platelet counts. Reference Mahawish, Pocock, Mangarai and Sharma35,Reference Mihalov and Timarova37,Reference Zhao, Lian, Zhang, Xie, Gao and Wang38 Furthermore, thrombopoietin receptor agonists can be given to maintain the level of platelets. The therapeutic dose of anticoagulation can be considered for ITP patients with no hemorrhage, stable hemoglobin = WHO grade 0/I/II and platelet count ≥50 × 109/l. There is no specific data for direct oral anticoagulation (DOAC) in this context.
Conclusion
We presented two patients with ITP and acute ischemic strokes who were treated with thrombolysis and EVT. In the literature, there are 27 patients with ITP who presented with acute ischemic stroke, and only one of them was treated with acute reperfusion therapy. In addition, there are 92 patients with significant thrombocytopenia with no available data regarding the cause of thrombocytopenia, who were acutely treated (tPA = 71, EVT = 21). Seven patients developed sICH (tPA = 5, EVT = 2), and one patient underwent EVT developed mild SAH. A conclusion cannot be drawn based on these limited number of published cases and with the lack of detailed information about the etiology of the thrombocytopenia. Treatment of acute ischemic stroke in patients with ITP requires close collaboration between hematology and vascular neurology experts to find a balance between the benefit and risk of hemorrhagic complications. Guidelines are lacking for acute stroke and stroke prevention management in ITP patients. Given the pathology of ITP thrombocytopenia is unique, patients with ITP cannot simply be included in groups of patients with thrombocytopenia due to other causes. Future stroke studies, such as multicenter prospective cohort study, that include ITP patients and patients with thrombocytopenia are needed to provide evidence-based treatment plans.
Acknowledgment
AA thanks King Saud University and the Saudi Arabian Ministry of Education for Residency and Fellowship funding. AA thanks The University of Alberta Hospital Foundation and the Neuroscience and Mental Health Institute for the Neurology Fellowship Award.
Disclosures
KB reports grants and personal fees from Pfizer, personal fees from Boehringer-Ingleheim, grants and personal fees from Servier, personal fees from Medtronic, outside the submitted work. CW is a member of advisory boards and honoraria: Pfizer, BMS-Pfizer, Leo-Pharma, Servier. Dr. Wu is a local Principal Investigator for trials with funding from: Bayer, Daiichi-Sankyo, CIHR, Heart and Stroke Foundation of Canada. The other authors have no conflicts of interest to declare.
Statement of Authorship
AA conceived of the idea, acquired the data, contributed to the medical management of the patients, and drafted the initial manuscript. KP, MA, GA, CW, and HLS made critical revisions of the manuscript. GB, KB, and JR contributed to the medical management of the patients and made critical revisions of the manuscript. KK conceived of the idea and made critical revisions of the manuscript.
