Indian Pediatrics 2002; 39:468-472
Spontaneous Intracranial Hemorrhage in Children with Immune Thrombocytopenic Purpura
Laxman Singh Arya
Immune (idiopathic) thrombocytopenic purpura (ITP), an autoimmune disorder characterized by antibody mediated destruction of platelets, is generally a self-limiting benign disorder with a 60-80% chance of spontaneous recovery occurring usually within a few months after onset(1-4). However, the risk of hemorrhagic manifestations, especially intracranial hemorrhage, prompts many physicians to consider some form of therapy. Although intracranial hemorrhage (ICH) is an extremely rare complication, it accounts for most of the mortality associated with ITP(3,5,6). Approximately 15-20% of the patients with acute ITP will develop a chronic form of ITP in which the thrombocytopenia persists for more than 6 months or a recurrent form in which the platelet count decreases after having normalized for at least 8 to 12 weeks(7). The risk of ICH is considered to dwindle after the first few days of presentation of the acute form and is exceptionally rare in chronic ITP(5,8). In view of absence of medical literature on ICH complicating ITP from the Indian subcontinent, our experience of ICH occurring in 8 of 240 children with ITP, over a 15 years period, encouraged us to present and compare the results of our retrospective analysis with the currently established concepts concerning care of children with ITP.
Subjects and Methods
The case records of all children, with a diagnosis of ITP, attending the Pediatric Hematology and Oncology Clinic at the All India Institute of Medical Sciences, New Delhi between January 1982 and December 1996, were reviewed for occurrence of ICH. The patients had fulfilled previously reported diagnostic criteria for ITP including isolated thrombocytopenia (<100,000/mm3) and normal or increased number of megakaryocytes in otherwise normal bone marrow aspirate(9). Besides a complete blood count with examination of a peripheral smear, the diagnostic work up included a thorough history and physical examination to search for any risk factors associated with ICH, as well as, a search for any clinical evidences suggesting other causes of thrombocytopenia like collagen vascular disorders, human immunodeficiency virus (HIV) infection, specific congenital syndromes and co-morbid conditions associated with a bleeding diathesis. For each child with ITP complicated by ICH, we obtained information regarding age, sex, type of ITP, platelet count at initial visit, platelet count on the day ICH occurred, interval from diagnosis of ITP to development of ICH, treatment received in the past, management during admission for ICH and outcome of ICH.
Children with acute ITP were treated with prednisolone (2 mg/kg) for 2 weeks followed by tapering over the next week. Children with chronic ITP were managed with either low dose corticosteroids, androgens or immuno-suppressive drugs. Besides ensuring a regular follow up, parents were advised to bring the child to the emergency department in the eventuality of any serious bleeding episode. Symptoms and signs of impending or evident ICH, in the form of headache, vomiting, convulsions, focal neurological signs and alteration of consciousness were treated with intravenous immunoglobulin, (IVIG) (1 g/kg for 2 days), intravenous methylprednisolone (30 mg/kg for 3 days) or intravenous dexa-methasone (0.5 mg/kg for 3 days). Computed tomography (CT) of head was performed for localizing the ICH.
Eight (3.3%) children with ICH (5 boys and 3 girls) were identified among 240 children receiving medical care services for ITP during the study period (Table I). Six of these children had chronic ITP at the time of presentation with ICH. Two children presented within 3 months of diagnosis of ITP. The mean age at presentation was 10.0 years (range, 0.75-17.0 years). The mean interval from the onset of ITP to the development of ICH was 1.9 months (range, 3 weeks to 3 months) in the two cases of acute ITP and 3.1 years (range, 1.25 to 6 years) among children with chronic ITP. The mean platelet count at the time where ICH occurred was 20,000/mm3 in acute ITP (n = 2) and 18,666/mm3 (range, 10,000 - 40,000/mm3) in chronic ITP (n = 6). Overall, the mean platelet count in both acute and chronic ITP with ICH was 19,000/mm3. Except for a single child, all patients had presented with symptoms of headache, vomiting, and drowsiness of more than 3 days duration, before arriving at our emergency department. None of the patients had any apparent risk factors like history of recent trauma, immunization, viral illness or drug intake. Computed tomography (CT) of the head was performed in 7 children and revealed large intracerebral hematomas in parietal lobes (n = 4) and frontal lobes (n = 3); none had any evidence of infratentorial bleed. Four patients were treated with intravenous dexamethasone, two with IVIG and one with intravenous methylprednisolone. No patient underwent an emergency splenectomy for control of ICH. One child survived after two partial seizures and is keeping good health for the past nine years till date. One boy (patient 5, Table ) with rapidly deteriorating neurological status, refused any form of treatment and left the hospital against medical advice, in a state of deep coma. Six of 7 children (87%) died within 24 hours of hospitalization.Table–Profile of Intracranial Hemorrhage in 8 Patients of Childhood ITP
Idiopathic thrombocytopenic purpura is the most common acquired bleeding disorder in children with a reported frequency of 4-8/100,000 children per year(4). ICH, an extremely rare but major cause of death in ITP has historically been reported to have an incidence of 1% or greater, but, recent studies indicate a much lower rate of 0.1% to 0.5%(4). Relatively few benign disorders in pediatrics have generated so much controversy as the management of ITP. Although the process of spontaneous recovery may be accelerated by administration of high dose corticosteroids or IVIG, the response is often only transient and does not clearly provide protection against the rare complication of ICH. Few studies have examined ICH in children with ITP. Information concerning the frequency of ICH in acute and chronic ITP, as well as risk factors leading to its suspicion or occurrence, are areas of current research questions geared towards a better understanding of its pathogenesis.
In a study by Medeiros and Buchanan(4), a world literature review between 1975 to 1996, revealed that only one of 56 children with ITP complicated by an ICH had a platelet count of more than 20,000/mm3, seventy three per cent of patients had a platelet count of less than 10,000/mm3. Although there are no definite predictors of which child will suffer an ICH, some factors like history of recent trauma, immunization, viral infection, menstruation, drug ingestion, adolescents with ITP, systemic lupus erythematosis, retinal hemorrhage, "wet purpura" including other severe muco-cutaneous bleeding and even cerebral arterio-venous malformation may possibly be high risk factors predicting ICH(4,9,10). A review of our experience, revealed a platelet count of more than 20,000/mm3 in 3 (37.5%) children of our study group. Interestingly, a recent report from the Japanese study group on ICH in childhood ITP also disclosed 3 out of their 8 patients having a platelet count of more than 20,000/mm3; there was no correlation with either the severity of bleeding symptoms or the platelet count at onset of ICH(9). Although we could not identify any inciting factors for ICH in our patients, the apparently well maintained hemostatic control in many patients with thrombocytopenia of ITP for yet unknown reasons makes the inquiry into factors predisposing to ICH a vexing issue.
Although there has been a consensus that ICH usually occurs within the first month after the onset of ITP, at present there is no evidence to suggest that there is a specific time period during which patients are at most risk of ICH and would benefit from treatment to increase the platelet count(4). Also, ICH may develop in children of any age, with older children and adolescents not necessarily having high risk. In the study by Medeiros and Buchanan(4), of the 51 children with ICH in whom duration of thrombocytopenia was reported, 26 (51%) presented with ICH within 4 weeks following diagnosis of ITP and the rest developed ICH between 4 weeks and 9 years (mean 27 weeks) of diagnosis. In the present study, 6 out of the 8 children with ICH presented more than 6 months after the diagnosis of ITP. Lilleyman(11), in his survey of ICH complicating ITP, found that 4 out of 14 (28.5%) children presented after more than 12 weeks of the diagnosis of ITP. These studies and our experience may provide some solace due to the widely prevalent fear of ICH occurring in the "early period" which provokes many physicians to treat in this period of presumed risk. Unfortunately, since ICH is so rare, controlled studies defining effects of treatment on the development of ICH cannot be performed easily on a scale which will yield statistically significant results(4).
Two currently established features of ITP needing appraisal in this study include, firstly, the finding that once ICH occurs, it should not be regarded as fatal (as more than 50% have survived in literature) and secondly, children who are at risk of developing severe hemorrhage have not been proven to respond to IVIG or corticoteroids(4). In this context, our experience with mortality occurring in 6 (one patient left against medical advice in deep coma) out of our 7 cases has been alarming. The 53.5% survival in an earlier report(4) was achieved by aggressive therapy including splenectomy (52%) and craniotomies (36%); the ICH occurred despite prior or concomitant therapy with corticosteroids or IVIG. The Japanese study group(9) also reported a reduction in mortality rates in their study of ITP; besides emergency splenectomy and neurosurgical treatment, splenic embolization and plasma exchange were also utilized as life saving measures in their patients with ICH.
In patients with ITP complicated by ICH, many authors recommend emergency splenectomy and in case of progressive neurological deterioration or posterior fossa hemorrhage, an emergency craniotomy(3-5, 10,12). After splenectomy, most patients show a rapid and sustained rise in platelet count and during this period a spontaneous resolution can be expected(3,5,12). Although our patients received optimum therapy in the form of platelet transfusions, high dose steroid treatment/IVIG along with supportive care, splenectomy and neurosurgical interventions were not performed. Factors including delayed medical help seeking, poor outreach hospital services and a rapidly deteriorating clinical progression may be some of the reasons leading to the situation in most instances where we had virtually no time to arrange for a splenectomy or a neurosurgical intervention. We subsequently do feel that the role of emergency splenectomy and craniotomy should be explored more energetically in our patients.
Our experience of ICH in patients with chronic ITP (n = 6) and in patients with counts above 20,000/mm3 (n = 3) may lend support to the advocates critical of the practice involving chasing of the platelet counts, especially in acute ITP; chronic ITP patients can also be suspected as a group likely to suffer form ICH with a presumably higher risk than previously conceived. In the future, large-scale prospective randomized studies, focusing on possible factors in the development of ICH and examining outcome variables, in addition to platelet counts, such as costs and side effects of therapy, as well as overall quality of life of the patient, will be needed to generate the best treatment strategy for each patient.
Contributors: LSA conceptualized and co-ordinated the study as well as contributed to drafting of the manuscript; he will act as guarantor for the paper. VPC helped in drafting of the manuscript. YJ collected the data and helped in drafting of the manuscript. HPP helped in clinical and hematological work up of the patients. SN helped in revision of the final manuscript.
Competing interests: None stated.