Portal de Periódicos da CAPES

Medical nemesis and childhood ITP

2003; Wiley; Volume: 123; Issue: 4 Linguagem: Inglês

10.1046/j.1365-2141.2003.04656.x

1365-2141

John Lilleyman,

Blood disorders and treatments

Thirty years ago the philosopher and theologian Ivan Illich accused the medical establishment of being a major threat to health. He argued that it was responsible for a growing epidemic of iatrogenic morbidity, and referred to this as a form of medical nemesis (Illich, 1974). He would doubtless say that little has changed, and, as far as childhood idiopathic thrombocytopenic purpura (ITP) is concerned, he would be right. I believe that doctors often make matters worse in the way they manage this condition. They prescribe mostly unpleasant, frequently unnecessary and sometimes dangerous therapy. They also offer self-defensive and over-cautious advice on risk management, creating what amounts to an anxiety state in some parents. To support these allegations, in this article I take a critical look at the real risk of dying from ITP, consider its lack of morbidity, and reflect on how little the multitude of plied therapies over the last century has actually achieved. I also suggest some organizational changes to improve the way affected children are managed in future. Childhood ITP is a diagnosis of exclusion and has become increasingly narrowly defined as most of the more serious conditions that used to fall under its umbrella have been weeded out. A century ago the terms ‘purpura simplex’ and ‘purpura haemorrhagica’ were uncritically applied to many symptomatic thrombocytpenias. These would certainly have included some leukaemias and aplastic anaemias because marrow examination in children was not routinely available until the 1950s (Wolman & Dickstein, 1947). Bundling more serious diseases under the ITP label, of course, gives a false impression of its mortality, as is clear from a description of 43 Manchester patients reported 50 years ago (Komrower & Watson, 1954). No fewer than six died (14%), but in some fatal cases the count did not fall below 90 × 109/l, so they clearly did not have uncomplicated ITP. Most of the children who perished probably had primary marrow failure, platelet functional disorders, immunodeficiency or malignant disease. As time passed, things improved. Of 180 Newcastle children reported 30 years later, only one died (0·5%), and based on the autopsy findings the authors felt that this was not because of simple ITP (Walker & Walker, 1984). Since then there has been a number of cohort studies where the recorded mortality has been of this sort of order or less, culminating in the largest to date from the Intercontinental Childhood ITP Study Group describing 2000 children from 38 countries where there were no recorded deaths at all, admittedly with incomplete follow-up, but also despite a selection bias towards more refractory disease (Kühne et al, 2001). The evident fall in mortality over the years is entirely explicable by the exclusion of the more sinister causes of thrombocytopenia. In the few sketchy reports of children who have apparently died of ITP, the terminal events are mostly complicated or raise questions about the diagnosis, and it should be appreciated that even today a significant number of children initially thought to have ITP turn out to have something else such as multi-system autoimmune disease or an inherited platelet disorder (Bader-Meunier et al, 2003). Intracranial haemorrhage (ICH) is effectively the only fatal complication, but even massive cerebral bleeds can be managed successfully with appropriate emergency measures and are by no means invariably lethal (Woerner et al, 1981). The true incidence is not known, but is unlikely to be anything like as high as the widely quoted 1%, even if children remain continuously and profoundly thrombocytopenic for months. In the first 2–3 weeks after diagnosis, the period of maximum anxiety for paediatricians, the figure is obviously much lower and was calculated at around 0·1% in a UK survey a few years ago (Lilleyman, 1994). Even this may be an overestimate, and no early cerebral bleeds were documented in the 2000 children in the Intercontinental Study – although two later bleeds were reported at the 6-month follow-up (Kühne et al, 2001). One child bled after 4 months and survived without sequelae, but the time of the event and the fate of the other child is not known. All this means that, despite paediatricians’ worst fears, the actual chance of dying from uncomplicated ITP around the time of presentation is tiny, and even in chronic unremitting disease is very small. The precise risk cannot be measured in the absence of better epidemiological and diagnostic data, but it is certainly small enough to live with and probably less, for example, than the risk of dying from epilepsy or asthma. So if the disease is so rarely fatal, in what other ways is it serious? ITP causes remarkably little pain or disability, although extensive bruising can be embarrassing and unsightly. Menorrhagia can be troublesome, nosebleeds can be a nuisance, and rarely gastro-intestinal bleeding can be a problem. On the whole, although, children who have ITP are perfectly well with or without a few bruises and petechiae. The worst problem in those whose thrombocytopenia fails to resolve quickly is often the proxy effect of parental anxiety. This is manifest by overprotection because of an exaggerated fear of fatal bleeding. Teachers and others who may be involved in the child's welfare can also become very nervous. The problem is clinically important because it can, of itself, become the main reason for therapy. And if it leads to splenectomy, the ultimate paradox can arise where the actual risk of sudden death has been increased but the fear of it has been reduced. Doctors must, of course, take most of the blame for this. They too often forget that life is for living and that children with chronic ITP are usually robust enough to take its bumps and scrapes with just a few more bruises than their peers. But paediatricians and haematologists are terrified of being sued for failing to warn of every eventuality, however remote or theoretical. They therefore scare parents with the prospect of a fatal stroke at any time, and tell them that children's activity must be restricted, that they should not be allowed to fly, must be supervised at all times, and should wear crash helmets. They then prescribe treatment to raise the platelet count and remove the perceived risk. But if mortality is so low and morbidity mostly so trivial, what is the potential benefit apart from reducing anxiety produced by medical advice? Illich would certainly call this iatrogenesis. Splenectomy was hailed as the treatment of choice in the 1920s (Williamson, 1926) 30 years before it was recognized to be dangerous (King & Shumacker, 1952). Then adrenocorticotropic hormone (ACTH) and cortisone were recommended (Komrower & Watson, 1954), followed, another 30 years later, by both polyvalent and anti-D intravenous immunoglobulin (IVIg) (Imbach et al, 1985; Becker et al, 1986). A number of other novel therapies were reported in the 1970s and 1980s but never achieved widespread popularity. Several management guidelines have appeared over the last few years (Eden & Lilleyman, 1992; George et al, 1996; British Committee for Standards in Haematology General Haematology Task Force, 2003), and from these two things are evident. First, little therapeutic progress has been made and secondly, the amount of grade C evidence hugely outweighs the grade A. In other words the advice is long on opinion and short on data. The lack of large randomized clinical trials compared with (for example) childhood leukaemia is depressing. The chief reason seems to be that care of children with ITP is not centralized, so trials have therefore been small with the largest to date recruiting only 146 children (Blanchette et al, 1994). This and the few others that have been achieved have nearly all looked at treatment for newly diagnosed children, and what they have collectively shown is quite clear without the need for a formal meta-analysis. Most patients will show a speedier rise in platelet count if given immunoglobulin (choice of doses and polyvalent/anti-D) than if given steroids (choice of doses and compounds), and a speedier rise with steroids than with nothing. What is not at all clear, however, is whether this matters in the slightest. Statistical significance is not the same as clinical importance and the platelet count in ITP is a poor predictor of morbidity (Sutor, 1998). As far as mortality is concerned, it is a sad irony that the only fatality recorded in the trials was an 11-year-old girl who died after 6 d despite having received both immunoglobulin and, later, steroids, and who had a ‘florid viral infection’ so may not have had simple ITP at all (Imbach et al, 1985). It is also relevant to note that the 2001 Intercontinental Study concluded that the outcome for children with ITP is entirely unaffected by initial treatment strategy (Kühne et al, 2001). This sits uncomfortably alongside a recent North American survey which showed that around 65% of such children would nevertheless be given some form of IVIg and 15% would be given steroids (Vesely et al, 2003). The few unlucky patients who progress to chronic unremitting ITP for more than 12 months, commonly overestimated in number and less than 5% of the total (Blanchette, 2002), present the most difficult management problems. The only randomized clinical trial of which I am aware recruited just 23 children (Hedlund-Treutiger et al, 2003), otherwise the recorded experience of treating chronic ITP consists of anecdotal reports and small observational studies, and all have to be assessed against a background of a disease that has a spontaneous remission rate stretching over several years (Reid, 1994). Various protocols of IVIg, anti-D, steroids, splenectomy and occasionally other more adventurous therapies have been tried (Blanchette et al, 1998). These, in addition to raising the platelet count in some of the patients for some of the time, have also produced fatal infection, suppressed immunity, haemolytic anaemia, blood-borne virus transmission, obesity, headaches, hyperactivity, peripheral neuropathy, hirsutism and hypertension. None have cured the disease in the sense of ending the underlying aberrent immune response; all merely produce a shift in the kinetics of the pathological processes involved pending a spontaneous remission. A possible exception to this generalization is rituximab, a chimeric monoclonal antibody directed against B-lymphocytes. This has produced interesting preliminary results in adults (Stasi et al, 2001), but experience in children is limited and the long-term effects, if any, are not yet known. So, what to conclude? Outside the context of emergency treatment for life threatening bleeding (which most paediatricians will never have to deal with during their practising lifetime) the effect of therapy on the mortality of childhood ITP remains to be determined but is probably close to zero. For morbidity, some measures, including splenectomy and perhaps rituximab, might improve the quality of life for the occasional child with chronic symptomatic thrombocytopenia by relieving disabling menorrhagia or self-consciousness, or by removing the straitjacket of fear. But heroic attempts to raise the platelet count in otherwise well children seldom justify the risks involved. Parents are still too often taken through despair, hope, confusion, and back to despair because they do not understand what ITP is or what is going to happen to their child, but in the UK the ITP Support Association is doing much to improve matters (http://www.itpsupport.org.uk). Through its website, seminars and newsletters, parents now feel less lonely, know more about the disease, and are taking a greater part in deciding how to handle their child's problems. This, in turn, has begun to change medical practice by educating paediatricians, and consequently there is a clear trend away from the must-do-something approach towards simply providing moral support while the disease burns itself out. The best way to manage the rare children with chronic, severe and unremitting ITP will never become clear while patients are cared for individually in local hospitals. What is needed is centralization of care in a few specialist centres – as occurs for other childhood blood diseases. That way, experience can be shared, better clinical and demographic data can be collected, and clinical staff in the referral centres can become more familiar with the disease. It works for haemophilia, it works for leukaemia, and there is no reason why it would not work for ITP. National referral centres could then set up National Registries and collaborate internationally to strengthen the laudable and still evolving Intercontinental Childhood ITP Study Group (Kühne et al, 2001). Demographic and epidemiological data could be improved and extended, and more multinational clinical trials carried out for defined patient sub-groups. Most importantly, risks could be better defined and we could then perhaps stop terrifying parents and start helping them to cope with their child's disease. Medical nemesis can be avoided if we try.