Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
The continuing debate about cardiotoxicity of antiopsychotics
- Zaubia Alyas (20 May 2001)
-
Re: The continuing debate about cardiotoxicity of antiopsychotics
- Marco Procopio (23 May 2001)
|
|
|||
|
Zaubia Alyas, SpR General Adult Psychiatry Homerton Hospital , London , E9
Send response to journal:
|
The Continuing Debate about Cardiotoxicity of Antipsychotics Coulter et al’s use of Bayesian methods is a commendable application of a complex statistical process for the identification of potentially risky drugs and adds to fuel the current debate that antipsychotic drugs themselves are important factors in sudden unexpected deaths occuring in psychiatric populations This concern was sparked by Reilly’s paper in the Lancet last year [2] , the findings of which culminated in the CoSM recommendation that the use of thioridazine in the UK be restricted exclusively to 1st line schizophrenics deemed not at risk of cardiac disease and curtailed the availability of anti - schizophrenic Droperidol to a named patient only basis . These changes have led to altered management strategies being adopted for various vulnerable psychiatric populations , the acutely psychotic , elderly and learning disabled people.Some would argue these changes have been for the worse but others would say it has perhaps led to more thoughtful prescribing with greater consideration given to the indication for prescription and less corruption of drug use as blanket cover. It is hard to know how tangible the conclusions of this paper are as they are based on probablilties that become less probable as they are extrapolated further away from the original data and in this case the original data is based on very small numbers just over 2000 cases drawn from 2 ½ million reported ADRs . Obviously many confounders may also come into play , for example : Clozapine is an end stage treatment .In this country it is only licensed for those defined as having treatment resistant schizophrenia and most people on it have previously undergone several lengthy trials of other classes of antipsychotics .Cardiac muscle changes , if they occur, may well be a chronic non - specific cummultative effect of the previous drugs rather than specific to treatment with clozapine . On this note , it may be informative to look at how early in the course of initiating clozapine treatment the diseases developed in a further retrospective study of the diseased cases. It would also be helpful to look at other cardiac disease risk factors in cases .The psychiatric population on clozapine are notorious for including those with the most coronary disease - prone lifestyle including people with chronic nicotine dependence , poor diet , sedentary lifestyles and mild to moderate obesity .It may also be useful to enquire about other inherited predispositions to cardiac disease in these groups Also Resperidone , haloperidol and lithium come 2nd , 3rd and 4th respectively after clozapine in terms of number of overall reports of ADRs so it is highly likely that selective reporting based on biased prescribing habits may also be acting as a confounder. This finding also makes me wonder whether antipsychotic induced myocarditis and cardiomyopathy could cause prolongation of the QT interval in some cases . From my basic knowledge of cardiac anatomy , there are 2 sorts of cardiac muscle tissue , contractile and conductive ,presumably both types could be affected by these patholgies thereby affecting conduction .However, the 2 diseases mentioned haven’t previously been implicated in cases with QT prolongation - various biochemical reasons have been cited as explanation instead. I am not confident that PM studies done on psychiatric patients dying suddenly and unexpectedly are sensitive enough to pick up early indicators of these heart diseases and even if they were pre - death ECGs aren’t always available to link the two [3] . References: 1. Antipsychotic drugs and Heart muscle disorder in international pharmacovigilance : data mining study , Coulter et al , BMJ , May 2001 2. QTc interval abnormalities and psychotropic drug therapy in psychiatric patients , Reilly et al [Newcastle], Lancet ,2000 3. Neuroleptic induced sudden death - case report and critical review , Boyd et al , [Ashworth Special Hospital] , Med Sci Law ,1995 Dr Zaubia Alyas
|
|||
|
|
|||
|
Marco Procopio, CONSULTANT PSYCHIATRIST LINWOOD CMHC, PRINCESS ROYAL HOSPITAL, LEWES ROAD, HAYWARD HEATH, WEST SUSSEX, RH16 4EX
Send response to journal:
|
Coulter et al's elegant analysis of a big WHO dataset offers several opportunities for discussion. The authors use bayesian statistics implemented in a neural network architecture and use in their analysis the following data:case reports for drugs,case reports of adverse reactions,reports of the specific combination,total number of reports. The authors' attempt to establish a correlation between the use of antipsychotic medication and heart muscle disorder using the above data is, on my opinion, non reliable. The possibility of having meaningful results from the authors' analysis of the above data is based on the assumption that the proportion between the total number of patients exposed to a certain drug and the total number of case reports of adverse reaction to that substance reported to the WHO centre is a constant similar for all drugs. It is intuitive that the reality is bound to be very different. The only way to have a meaningful result would be to include in the analysis the number of prescriptions for each drug during the period analysed in the 60 countries object of the study. The authors' analysis is otherwise bound to overestimate some adverse reports for some drugs and underestimate others. Another issue is that to be able to generalise results in bayesian statistics using a neural network it is indespensable a thorough knowledge of the subject at hand. I hope that the authors will forgive me for saying that the presence of a psychiatrist amongst them would have been extremely useful. An example is the fact that the overepresantation of case reports for adverse events for clozapine will not surprise any clinician who prescribes regularly that medication. As mentioned in the article there is a substantial risk of agranulocitosis amongst the patients who use clozapine. For this reason monitoring services have been developed in several countries and the clinicians are very attentive to any adverse reactions in these patients. As a consequence the number of reports of adverse reactions is inflated in comparison with the other drugs. The unique position of clozapine therefore does not allow generalisations and comparisons with other drugs in the framework set by the authors. I am very surprised that lithium is classified in the article as an antipsychotic medication. Clinically lithium is a mood stabiliser and not an antipsychotic. From the chemical point of view is one of the smallest ions while the antipsychotics are complex organic molecules. From the pharmacodynamic stand lithium acts on the prostaglandin methabolism while antipsychotics seem to function through a blockade of the dopaminergic transmission. If the data on lithium are excluded from the group named "other antipsychotics", as it should be done,the association between these drugs and heart muscle disorder becomes even weaker. The authors conclude that the results could have been obtained using a simpler method, but that all simpler methods rely on someone looking for an association. I personally look at this characteristic of simpler methods as a strenght and not as a weakness:it is called developing an hypothesis, the cornerstone of all proper scientific methodology. |
|||