HIV TREATMENT AS PREVENTIONPosted on March 27th, 2009 2 comments
The proposal that testing and treating everyone who is HIV infected would end the epidemic is back in the news.
It is not a new idea. It has been discussed at HIV/AIDS conferences. At the beginning of the year an exercise in mathematical modelling was presented in the Lancet providing some support for this notion of universal testing and treatment. Now some experts in molecular biology and virology have added their personal opinions in favour of this approach; I notice that at least on one web site addressed to HIV and Hepatitis virus infected individuals, the views of these pioneer researchers are reported with, as seems to be usual, no analysis or criticism. http://www.hivandhepatitis.com/recent/2009/032409_a.html
It begins to look almost like an advertising campaign, with the touch of a skilled publicist; an idea is gradually brought to public attention, it is widely endorsed and the hope is that public support will ensure that funders and politicians will move the project forwards.
The merits of the proposal, and the way it is being promoted are two different issues.
Regarding the proposal, in principle it is certainly a worthwhile idea that deserves consideration.
But there are several problems, not mentioned in public reports of this proposal, and barely dealt with in the professional literature, which may constitute insuperable barriers to its implementation.
Leaving aside for the moment the question of whether such a project is even feasible, perhaps the most important problem is that infected people who do not need treatment will be asked to receive it to achieve a social benefit.
This proposal then involves the general concept of a public health intervention on individuals who will not themselves derive any benefit from the intervention, but will only be exposed to its risks.
We have thankfully not yet reached the point where enforced testing and treatment can be seriously proposed. (We may be getting close in the removal of written informed consent for HIV testing).
Certainly the spectre of mandatory testing and treatment is lurking behind this proposal to test and treated everyone infected. This would do wonders for drug and testing equipment sales.
So we would have the situation where some individuals will voluntarily take treatments that despite what we may be told can most certainly not be regarded as absolutely free of possible adverse effects, Many infected people will of course benefit from this. Others however will agree to take risks, with no benefit to themselves but for the benefit of others. Quite apart from many other issues, we can only ask these individuals to participate in the project if there is an overwhelming chance of success. At the moment we do not have this assurance.
It is not a digression to compare this situation with that in which an individual is asked to join a clinical trial and who may be randomly assigned to receive a new treatment of as yet only conjectural benefit. We are absolutely obliged to ensure that the trial design is such that reliable information will be obtained from the study.
Since the testing and treatment of all infected individuals to end the epidemic can in no way be regarded as an undertaking with an assured successful outcome, it really is a trial, based on an hypothesis somewhat supported perhaps by mathematical modelling. As such it will require written informed consent from the participants.
I wonder what such a consent form would look like. It is possible, actually likely, that a consent form outlining possible risks and benefits would dissuade many from participating.
The disincentive would be felt by those infected individuals who do not themselves require antiretroviral treatment.
This inconvenient obstacle can be easily eliminated.
All that is needed is for treatment guidelines to include a recommendation that antiviral treatment should be offered to all infected individuals, even those with greater than 500CD4 lymphocytes. A precedent has now been set where treatment recommendations can be made on the flimsiest of evidence. The inappropriate use of retrospective observations to justify an earlier start to antiretroviral treatment is a good example.
So all one needs to do is to move the goal posts a little further and declare that antiretroviral treatments should be given to all HIV infected individuals, irrespective of CD4 count. There should be no difficulty in selecting retrospective observations that will support this recommendation. In the field of HIV/AIDS you can probably find retrospective data to fit whatever idea you are interested in promoting.
There is another tool available to promote the contention that every HIV infected individual, irrespective of CD4 count will benefit from antiviral therapy. This useful tool is called “expert opinion”. (Actually, people billed as “experts” have already expressed this opinion).
The problem with this is: what does it take to be regarded as an expert?
We may well be in an era where we have “experts” for hire.
Defining what was meant by “expert” was once much easier. Years of experience and significant contributions to the field might have been required attributes. But no longer.
Experts can seemingly be created overnight, at least by commercial entities interested in marketing a product. Their credentials are easily supplied. These instant experts will give talks at conferences, they will appear on educational programs, and even put their names to ghost written articles.
Revealed: how drug firms ‘hoodwink’ medical journals Pharmaceutical giants hire ghostwriters to produce articles – then put doctors’ names on them.
As for the practice of ghost writing , there is a great deal of evidence for this, a little shown above. I’m ashamed to admit that I once (only once many years ago) allowed an employee of a drug company to write an article which carried my name. But I had done the work without their support, and in my defense, I checked every word, changing some, – an experience the writer was evidently not used to. This was my first (and only) personal encounter with this practice
I will hazard a prediction; before the year is out we will have arrived at the point that experts will state that every HIV infected person benefits from treatment, irrespective of CD4 count. If required we will see retrospective observational studies which show that in people who started treatment above a CD4 count of 500, mortality from all causes was reduced as compared to those starting below 500 CD4 cells. It should be just as easy to find retrospective data that shows that starting treatment immediately on diagnosis confers a benefit not seen when treatment is delayed to CD4 count of 350.
Of course these expert views will be very widely disseminated in press reports and on numerous web sites – some will even provide the opportunity for doctors to earn CME credit. In this way conjectures are transformed into established facts.
I don’t know how we might obtain real evidence that testing and treating all infected people is not only feasible, but would achieve its goals. The two are related.
For example, how does one ensure that all people are tested? Or that they will agree to be treated? Or that they will adhere to their treatments?
As imperfect as this is maybe one approach is to test these issues in a limited setting where mobility in and out of the selected areas can be controlled for.
This could more usefully be a trial where two different strategies were compared – the present practice of starting treatment at 350 CD4 cells, and treating everyone infected, while promoting HIV tests in both groups. Despite complications introduced by the movement of people, we might get an idea if this is a feasible and effective approach.
Sadly those bodies that instruct physicians on how to treat HIV infected people, and who tell HIV infected people what is best for them, seem to be averse to calling for prospective studies, designed to shed some light on what may in fact be best for infected people. Those who manufacture the treatments appear to prefer trials that are designed to provide them with the answer most congenial to them. Here is an account of the practice of designing trials to provide the answer most desired.
They can also rummage in retrospective data collections selecting observations best suited to the outcome they have already decided on. Of course there is always an expert to be created to promote this outcome.
When the mathematical modeling referred to above supporting the idea of a “test and treat everyone infected” approach appeared, I wrote a reply to the Lancet which published the article. Not my letter, which was politely rejected.
I am adding a slightly edited copy of that letter here.
A recent Lancet article suggests that we could end the HIV epidemic by testing and treating all who are infected, irrespective of whether or not the individual would benefit from such treatment (R. Granich et al. 2009 Lancet 373:48).
This represents an intervention on individuals, primarily for a public health benefit. At the present time, ethical considerations make this proposal a completely indefensible approach.
The available drugs are far from benign; for a particular individual, their use is desirable and justified when their benefits clearly outweigh their risks. Treating individuals to achieve a population benefit requires a similar risk benefit assessment. F M Hodges and colleagues have addressed this issue. (EM Hodges, JS Svoboda, RS van Howe
Prophylactic interventions in children: balancing human rights with public health. J Med Ethics 2002; 28: 10-26)
To protect individual liberties they propose six conditions that should be met before for such interventions are taken. All of these are reasonable. I quote a passage from their article that outlines them.
“PROPHYLACTIC INTERVENTIONS FOR PUBLIC HEALTH BENEFIT”
Prophylactic medical interventions are frequently performed on healthy individuals who have given informed consent. …..
The most common example arises when the patient is at significant risk of contracting a life- and public health-threatening illness for which the proposed prophylaxis is a proven preventive. In order to safeguard individual liberties, the situations in which such procedures may be undertaken for public health benefit must meet the following requirements:
1. The danger to public health must be substantial.
2. The condition must have serious consequences if transmitted.
3. The effectiveness of the intervention in safeguarding the majority of the public against the particular malady must be well established.
4. The intervention must be the most appropriate, least invasive, and most conservative means of achieving the desired public health objective.
5. The individual must be provided with appreciable benefit not dependent on speculation about hypothetical future behaviours of the patient.
6. The burden to the individual’s human rights and health must be balanced against and found to be substantially outweighed by the benefit to society in helping prevent a highly contagious disease or other potentially calamitous condition from affecting the public health”.
Clearly the proposal to treat all infected people will include some in whom the fifth consideration will not be met, but the concerns are covered in the sixth one. But here the benefit to society must be assured, or more practically, be considered to be highly probable, with credible evidence produced to support the contention (as stated in the third consideration).
While the first two criteria are very clearly met, the present proposal to treat all who test positive fails utterly on the third point. It is far from well established that antiviral treatment of all who are infected will protect the “majority of individuals” in diverse settings. Among problems acknowledged by the authors are those related to toxicity, adherence and the development of resistance to the antiviral drugs. To this must be added the possible negative effects on behaviour deriving from a perception of being non infectious. The fourth condition is also not met. We cannot state that we have exhausted the utility of prevention education and promotion of condom use.
Let alone the questionable wisdom of mounting an extensive and expensive public health intervention that is based only on mathematical modeling, we are very far from possessing information that would supply the slightest confidence that such a measure would effectively meet its objective.
Regarding adherence, the optimism presented by the authors based on studies in Malawi is hardly justified. Adherence by individuals who may be ill, and certainly know they are receiving medications for their own benefit tells us nothing about adherence by people who feel healthy and know they are not taking the medications to benefit themselves.
The general relationship between viral load and infectivity is well established. The success of the proposed strategy according to the model presented depends on achieving a significant reduction in viral load from the pre-treatment value. The solid evidence of the potent ability of antiviral drugs to very substantially reduce viral loads in a sustained fashion derives predominantly from observations in settings where untreated endemic or concurrent infections are uncommon. The ability to achieve a sustained significant drop in viral load may be more difficult where there is a high prevalence of untreated endemic or associated infections. This is the case in parts of Sub Saharan Africa. Many of these infections are able to activate and enhance HIV replication, through the action of pro inflammatory cytokines. Should these infections be associated with genital ulceration there are additional uncertainties.
HIV disease is characterized by an enormous variability in the rates of disease progression. There is no such thing as a standard course of disease progression that is one of the assumptions used in the modeling. We know very little about the distribution of different rates of disease progression among infected individuals, or about the influence on this of associated untreated infections.
Risking individual harm for a public benefit is a slippery slope. Will we see a proposal to administer (with consent, of course) antiretroviral medication to the whole sexually active population, HIV infected or not?
AIDS is a preventable disease. We have far from exhausted less conjectural, as well as less speculative approaches to its prevention.
Apart from this proposed strategy to treat all infected people, there definitely are situations where treatment as prevention is absolutely appropriate and desirable. One is post exposure prophylaxis (PEP), where individuals who have been exposed to HIV attempt to prevent infection by rapidly taking antiretroviral drugs – that is within 72 hours of exposure. This applies to both occupational and sexual exposure. Regarding sexual exposure – where feasible, which is certainly the case in N America Europe and in many other regions, a 3 day supply of drugs should be available 24 hours of the day, given the limited time frame for action. Measures to immediately start PEP immediately should of course be available where occupational exposure is a risk. Emergency departments should be equipped and ready to start the protocols for PEP. People at risk should even be encouraged to keep a 3 day supply of drugs at home to cover times when medical care is not available – at night or weekends. .Very importantly people at risk must be informed of the availability of PEP.
The second is pre exposure prophylaxis. This is taking antiretroviral drugs on specific occasions when there might be a risk of exposure. This absolutely cannot replace the use of condoms, but some individuals may wish to take an additional even if unproven preventative measure. This really is a matter for individual choice. Our obligation is to make it very clear that this is not a substitute for condoms.
Leave a reply