Industry’s Influence

Do they or don’t they?

It’s the question that paper after paper after paper after paper has tried to address over the past decade: does industry sponsorship of clinical trials bias the outcomes of the study? Does such sponsorship influence how doctors interpret the research findings? Does it affect healthcare professionals’ willingness to believe the results and integrate an experimental treatment into their practice?

The latest attempt to navigate the thicket of pharma funding is in this week’s New England Journal of Medicine. In a paper titled, “A Randomized Study of How Physicians Interpret Research Funding Disclosures,” Aaron S. Kesselheim and co-authors report their results of a study designed to test how scientific rigor and funding disclosure influence the credibility of a clinical trial in the eyes of doctors.

(And let’s just get the obvious out of the way now: Kesselheim et al’s study was not funded by a pharmaceutical company. The disclosure forms are available online as a PDF.)

Survey says…

Here is how the authors set out to determine how funding disclosure affects what doctors make of a study’s findings. First, they made up studies for three new drugs for common disorders: hyperlipidemia, diabetes, and angina. And really, the invented drug names are spectacular: lampytinib, bondaglutaraz, and provasinab. Bondaglutaraz! Fantastic.

Then, after what must have been a very enjoyable name-creating session, the authors wrote fake abstracts describing fake trials. The abstracts varied according to their level of scientific rigor (high, middle, and low for each drug) and funding status (no funding source mentioned; funded by the NIH; or funded by a pharmaceutical company with a lead author that was financially involved with the sponsoring company). The companies fake-sponsoring the studies were selected at random from the 12 pharmaceutical companies on this list.

So – three drugs, three study designs, three funding types, for a total of 27 abstracts. Then it was time for the survey, which was mailed to 503 physicians randomly selected from a total 45,398 applicable physicians listed with the American Board of Internal Medicine. The physicians were offered $50 for completing the survey in several notifications. Nonrespondents were even mailed a crisp five dollar bill, with promise of the remaining $45 upon return of the completed survey.

The physicians weren’t asked to review all 27 abstracts. Rather, each participant was asked to review three abstracts, each for a different drug. The physicians knew that these were made-up drugs and they were told to assume that the drug had recently been approved by the FDA and was covered by insurance – in other words, satisfying all the criteria that would otherwise be needed for a doctor to consider a drug safe and practical to prescribe. The levels of rigor and funding were selected at random for each of the three studies provided to each physician.

The survey asked the respondents to score each abstract according to their likeliness to prescribe the drug.

What is a rigorous clinical trial?

It’s worth pausing here to talk about what makes a clinical trial rigorous. The question at the heart of this NEJM report is: If a trial is conducted in a highly rigorous manner, will industry funding still diminish its credibility? And the accompanying, more editorial question: Should industry funding diminish the credibility of a rigorous clinical trial?

Here is what makes a clinical trial rigorous (according to the NEJM paper authors, and just generally speaking):
• The study should randomized
• The study should be double-blind; that is, neither the investigators nor the participants know who’s receiving which treatment (the experimental or the control)
• The study should have an active comparator; that is, it should not be a single-arm study, but should be comparing two or more regimens
• The dropout rate (the number of enrolled patients who leave before the study is concluded) should be less than 9%
• For these hypothetical new drugs, the study was to have a sample size (ie, total study population) of 5,322
• The enrolled patient population should accurately represent patients with the disease/condition in question. In other words, if the average age of patients with the condition in question is, say, 65, then the average age of patients in the study should not be, say, 35.
• The drug should be documented as being safe
• For the hypothetical drugs, a rigorous study was one with a follow-up of 36 months; that is, patients had been treated for at least three years with the experimental drug

Okay, so. With the three abstracts in hand, the physicians were asked to say how likely they were to prescribe the new drug (1=very unlikely; 7=very likely), and to score the methodological rigor of the trial (1=not at all rigorous; 7=very rigorous), and to score their confidence in the conclusions reached by the investigators (you guessed it: 1=not confident; 7=very confident). And participants were also asked to respond to the juicy question:

“Do you think that pharmaceutical company funding is likely to influence the outcome of scientific studies about the efficacy and safety of pharmaceuticals in favor of the drug in question.”

(Survey respondents were also asked to disclose their own financial support received in the previous year.)

Dim the lights, it’s time for the results!

Well, despite the promise of fifty bucks, only about half of the solicited physicians responded to the survey (53.5% to be exact, for a total of 269 respondents). Interestingly, about 75% of the respondents reported receiving at least one type of industry support in the past year. And they generally agreed with that question about whether industry funding can influence trial outcomes.

Also – there was a high correlation between the rigor of the abstract and the perceived rigor. In other words, the respondents correctly discerned which abstracts were for the most rigorous studies and which for the least. The confidence in the study outcome matched the level of rigor; the least rigorous studies engendered the least confidence in the results.

More interesting, there was a definite, clear association between the funding disclosure and how physicians perceived the rigor and results of a trial. When industry funding was disclosed, the trial was perceived as less rigorous compared to abstracts where no funding was disclosed. And, regardless of the design, physicians had less confidence in the results of industry-sponsored trials, even when the trial was highly rigorous (I know, the word “rigor” is getting irritating right about now. Sorry for that. My inner thesaurus is on a coffee break right now). The respondents were also less willing to prescribe drugs tested in industry-funded trials compared to drugs tested in trials with no funding listed.

Also:
• Industry-funded trials were considered less important than NIH-funded trials
• Respondents were less interested in reading the entire report of an industry-funded study than they were in reading the entire report of an NIH-funded study
• The distaste for industry-funded research was evident at all levels of scientific rigor
• US-trained physicians were less likely to say they were willing to prescribe any of the pretend drugs than were physicians who’d trained outside the US.
• Older physicians were more likely to say they’d prescribe the new drugs than were younger physicians

As the authors write in their discussion:

“…respondents downgraded the credibility of industry-funded trials, as compared with the same trials randomly characterized as having NIH funding or having no source of support listed. The magnitude of this reduction in perceived methodologic rigor was about the same as that for low-rigor trials as compared with medium-rigor trials. Physicians’ skepticism of industry-funded research affected their responses to high-rigor and low-rigor trials similarly.”

The authors see the results of their survey as problematic. They point out that the pharmaceutical industry has funded the study of many major drugs that are now clinically important, and express their concern that excessive skepticism could hinder the translation of research findings into clinical practice. They call on the pharmaceutical industry to address this thinking so that the credibility of clinical trials will be more weighted in its rigor than in how its paid for. “The methodologic rigor of a trial, not its funding disclosure, should be a primary determinant of its credibility.”

The source of skepticism

The skepticism evident in the surveys has accrued over years. There was this lawsuit, in which GlaxoSmithKline was accused of concealing data. There’s the recent $322 million fine of Merck over off-label promotion of Vioxx, just a year after the company paid a $1 billion penalty for the way it promoted rofecoxib. Abbott recently paid $1.6 billion for promotion of unapproved uses of Depakote. And then there are the many disparaging insights in The Truth About the Drug Companies.

Yet in an editorial accompanying the report of the survey results, Jeffrey Drazen, MD, points out how many investigators in NIH-sponsored studies have incentives such as academic promotion and recognition that could influence the outcomes of their trials. He writes:

“A trial’s validity should ride on the study design, the quality of data-accrual and analytic processes, and the fairness of results reporting. Ideally, these factors – not the funding source – should be the criteria for deciding the clinical utility. Patients who put themselves at risk to provide these data earn our respect for their participation; we owe them the courtesy of believing the data produced from their efforts and acting on the findings so as to benefit other patients.”

That last clause might be a bit of a stretch. Feeling beholden to the patients who enrolled in the trial as justification for believing the results of the study? I’m not sure that I’d want my doctor prescribing me a drug because he owed it to the patients who’d been part of its investigation. But the call for balance seems warranted. After all, unless some very vast change sweeps across the twisted world of medical care, industry funding of clinical trials is here to stay.

Category: Drug Development, Housekeeping, Industry funding | Tagged , , , , , , , | Comments Off

The Battle Over Generic Drugs in India (and elsewhere)

A few days ago, in a courtroom in India, a landmark ruling was made that could impact future legal battles over the sale of generic versions of expensive medications in India. When the Delhi High Court rules that Cipla, the country’s largest drug maker, could sell its own version of the lung cancer drug, Tarceva (erlotinib), made by Roche. The generic version will reportedly cost about a third of the proprietary version.

The lawsuit had been brought by Roche against Cipla. Roche accused Cipla of patent infringement, but the judge, Justice Manmohan Singh, sided with Cipla, agreeing that the generic version has a different molecular structure from Roche’s original. The ruling upholds Roche’s patent on its molecular structure, but allows Cipla to produce its own version, which (presumably) has the same mechanism of action and yields the same outcomes as the Tarceva molecule. The case had been ongoing for four years and the ruling is considered to be a landmark judgment, though Roche could very well appeal.

The timings are interesting, as they seem to set the stage for the reopening of another years-long drama involving yet another cancer drug. The Switzerland-based pharmaceutical company Novartis sued the Indian government years ago over its decision to allow generic versions of the drug to be sold there. First filed six years ago, the case was reopened just yesterday. At the same time, Pharmalot reports that the Indian court has just told Novartis to lower the price at which Gleevec (Glivec outside of the U.S.) is sold in India.

The patent battle over Gleevec is, needless to say, bitter and complicated. Gleevec is a tyrosine kinase inhibitor that treats chronic myelogenous leukemia (CML), a rare cancer that is slow-growing but fatal. Gleevec targets a malfunctioning enzyme that is the product of a genetic abnormality present in more than 95% of patients with CML. For almost all patients with CML, Gleevec or its second- or third-generation counterparts enable people with CML to live a normal, healthy life (and lifespan) with minimal, if any, side effects. (Forgive this brief commercial interruption, but my forthcoming book, The Philadelphia Chromosome, tells the incredible story behind the creation of this groundbreaking drug. More on that another time!)

Gleevec is expensive, but somewhere in the neighborhood of 90% of CML patients in India get the drug for free. So either the company is still making a bunch of money from the remaining 10% of paying patients (some of whom receive a deep discount) or Novartis is fighting the ruling as a way to establish clear guidelines for future branded drugs that may be sold in India.

Although the drama between Novartis and the Indian government is a fascinating case study in its own right, the fight also raises larger questions about drug patents abroad. Cipla was established in the 1930s by Khwaja Abdul Hamied, a nationalist and follower of Mahatma Gandhi. The company became famous in 2001 when it decided to sell a triple-combination antiretroviral therapy for HIV/AIDS for a price that came to less than a dollar per day, a shock in light of the $10,000-per-year cost of antiretroviral therapy produced by other pharmaceutical companies.

Patents on pharmaceuticals are a contentious issue for many countries around the world, for the obvious reason that the average person in most countries can’t afford to pay for brand-name drugs. The high price of many patented drugs goes against the belief held by many governments that citizens should be guaranteed the best possible delivery of healthcare. In 1994, the World Trade Organization administered The Agreement on Trade Related Aspects of Intellectual Property, known more commonly as TRIPS. The agreement increased the scope of protection for intellectual property that has had the possibly unforeseen consequence of restricting access to medications among impoverished people.

Interestingly, when the discussions that eventually led to TRIPS first began in 1986, fifty countries did not provide patents for pharmaceutical products. Stretching back even further, India had implemented a policy in 1970 that permitted “reverse engineering,” a process used to create generic versions of medications in developing countries — essentially, the Indian government was disallowing patented drugs.

Currently, all WTO Members except for those categorized as “least developed countries” are required to issue 20 years of patent protection in all fields. The purpose was to spur innovation (isn’t that always the stated purpose behind patent regulations?) but the outcome has been restricted access because many people in WTO countries cannot afford to pay for the drug, and no one is allowed to sell a generic version.

Subsequent amendments have attempted to solve the problems resulting from TRIPS. There’s TRIPS flexibilities, the Doha Declaration of 2001, and measures taken by individual governments (for example, rulings in Brazil, Venezuela, and South Africa that ensured access to antiretroviral therapy).

Reverse engineering, the process that allowed Cipla to become one of the world’s most prominent generic drugmakers, has been ruled unlawful by India’s Patents Act, passed in 2005, and will no longer be permitted as of 2015. When that Act comes into play, it could mean the end of an era for generic drugs in India.

In the meantime, the patent battle over Gleevec in India continues, as do concerns about ever-diminishing access to medications worldwide. Once again, it’s the tussle between the fact that drug making is a for-profit business and the wish for unrestricted access to beneficial medications for people with serious illnesses everywhere. At the very least, people with lung cancer in India can look forward to generic Tarceva.

Category: cancer, drug patents, ethics, generic drugs, Healthcare disparities | Tagged , , , , , , , , , , , , , , | 2 Comments

Is Cancer Care Worse in Poorer Neighborhoods?

According to one recent study: no.

To be more precise, according to one recent study, areas of social deprivation have an equal supply of cancer care services as more well-to-do areas.

Elizabeth Lamont and colleagues, of Harvard Medical School, set out to determine “whether area social factors are associated with the area health care supply,” according to the abstract of their report, published in the Journal of Clinical Oncology. Using Census Bureau records, the authors noted the social factors at play in 3,096 urban zip codes, and checked the health care supply in the 465 hospital service areas that corresponded to those zip codes. Focusing specifically on breast cancer and colorectal cancer, the authors cross-checked the data to see if social factors were associated with the supply of health care services, namely screening, treatment, and post-treatment surveillance.

The authors found no such association. The supply of physicians conducting these three areas of care was no different for disadvantaged neighborhoods than it was for more advantaged neighborhoods.

In the quest to understand why health outcomes are worse among people living in poor areas, this study is a small step that rules out one potential variable. Okay, so, the number of people providing important services isn’t the problem. So if there is an adequate supply, then what’s the problem?

As Sandra Swain notes in a Reuters article on the study, insurance, money, medical literacy and transportation may all be contributing to the difference in outcomes.

Though the causes remain unclear, the link between social deprivation (is that a euphemism or a more nuanced and accurate way to speak about poverty?) and worse outcomes among people with cancer is undeniable. A few startling findings:

Among 7290 colorectal cancer patients in the UK who underwent surgery, social deprivation was associated with higher postoperative mortality (death after surgery) and a longer hospital stay — and that difference was on a sliding scale; as deprivation worsened, so did outcomes.

This report from Cancer Research UK notes that unskilled workers are more likely to die from cancer than are professionals, and cancer mortality rates, though varying widely across geographic areas in the country, are highest in areas with dramatic levels of social deprivation. Interestingly, the report authors assert that much of this difference can be chalked up to smoking prevalence.

But smoking can’t be the only factor. Another study of 762 women with breast cancer (also in England) found that wealthier women were less likely to be diagnosed with invasive ductal tumors, high-grade tumors, and estrogen receptor-negative tumors. That study also found that poorer women may have been having more potentially unnecessary mastectomies.

This report presents a more global view of the socioeconomic determinants of cancer, with an emphasis of the problem in developing countries.

The connection with social deprivation and race has also come under some scrutiny. As this report (albeit from several years ago) highlights, breast cancer incidence may not be that different among white women and black women, but breast cancer mortality is higher among the latter.

It’s true that most of the reports noted above are not from the United States, and perhaps the supply of services is an issue in other countries. My guess would be not, that the issues linking social deprivation to cancer in developed countries are, by and large, the same. But what are they?

Category: cancer | Tagged , , , , | Comments Off

(More about) Using Treatment to Prevent HIV

This entry was first posted as one of a series of posts by PLOS Network bloggers on Reuters.com.

Despite widespread knowledge of how the AIDS-causing virus HIV is transmitted, and how to prevent it, the disease is still spreading. An estimated 34 million people worldwide were living with HIV/AIDS in 2010, according to the World Health Organization. Sub-Saharan Africa is the worldwide epicenter, with 22.9 million people living with HIV/AIDS, but epidemics and areas of high concentration abound elsewhere, including in Western Europe and the southern United States. Now, an emerging concept known as treatment as prevention – where patients are given medication for the primary purpose of stopping new infections – is gaining favor as a way to decrease the spread of HIV, if not end it altogether.

Antiretroviral therapy (ART) can halt the progress of HIV in infected people, and prevent them from passing it on by reducing the amount of virus in the body. It is rarely initiated at the time of diagnosis because the disease usually isn’t causing physical deterioration at that point, but the lag between diagnosis and starting ART leaves a wide window for new infections. (WHO guidelines recommend starting ART when the level of CD4 cells – an infection-fighting component of the immune system that is gradually destroyed by the virus – plummets below 350 cells per cubic millimeter, far below the normal range of 500-1,500 cells per cubic millimeter).

For years, studies have explored whether starting ART earlier could help prevent the spread of HIV. A landmark finding published last year in the New England Journal of Medicine came from the large HPTN 052 clinical trial which found that starting ART early reduced the rate of transmission between stable, heterosexual partners by 96%.

But the concept of “treatment as prevention” is controversial and complicated. Should infection-preventing medicine be given to people who may not benefit from it themselves? Knowing that finances are limited, which individuals should be prioritized?

HPTN 052 focused on stable, heterosexual couples in which one partner was infected and the other not. Are those results applicable to the broad, diverse HIV-positive population? Considering the urgent need to thwart continued spread of the virus, how can these questions be answered in a timely, ethical manner?

A collection of articles just published by PLOS Medicine ahead of AIDS 2012, the 19th International AIDS Conference currently underway in Washington, offers an in-depth look at the problem, including mathematically modeling the cost and potential benefit of expanding access to ART. These models quantify: behaviors that lead to transmission, the biology of the infection, patterns of contact between infected and uninfected individuals, and the many other variables that make HIV so complex and difficult to study in actual populations.

A paper analyzing 12 distinct models finds that expanding ART access to 80% of the HIV-positive population, (and where 85% of those patients remain in care) could reduce the incidence of HIV by 35%-54% by 2020 (although the long-term implications are less clear).

Another looking at economic issues finds that simplifying treatment delivery – such as by not monitoring disease before ART begins, or eliminating CD4 count measurements – could offset the added expense of scaling up treatment-as-prevention efforts.

Wim Delva (University of Stellenbosch, South Africa) and colleagues note there are benefits and potential pitfalls with picking which groups of patients to prioritize. Targeting stable, heterosexual couples for treatment-as-prevention efforts makes sense in light of the HPTN 052 findings, but would be controversial among a population where such couples are relatively rare. Expanding ART access among female sex workers could dramatically reduce HIV transmission, though a lack of adherence to care may diminish that potential benefit. Prioritizing ART treatment according to viral load, rather than CD4 counts, may be a more effective approach for preventing new infections, though evidence supporting the benefit of this strategy for the infected patient is lacking.

Observations of real-world populations often contradict the findings of mathematical models, leaving researchers and policymakers perplexed about how to implement treatment as prevention strategies. The question of how to best use the tools that have been shown to reduce HIV transmission will likely dominate the field of HIV prevention for the foreseeable future.

Category: HIV | Tagged , , , , , , , | Comments Off

HIV Treatment as Prevention: Trying to Clear the Fog

Last year, a landmark clinical trial on the use of antiretroviral therapy (ART) to not only treat HIV but also prevent further transmission revealed that early initiation of ART reduced the rate of new infections among heterosexual couples by 96 percent. That study, HPTN 052, spurred numerous questions among healthcare researchers and policy makers. Should all patients with HIV start ART sooner after their diagnosis than has been the case thus far? If not everyone, what populations should be prioritized for this approach? What are the costs associated with earlier initiation of ART? Can patients who might not benefit clinically from earlier treatment be given ART strictly as a means of preventing new infections? Considering the sprawling range of HIV, how can these questions be examined quickly and accurately?

These and other complex questions are tackled in a new collection of articles recently published by PLoS Medicine. Entitled “Investigating the Impact of Treatment on New HIV Infections, the collection wades through the extraordinarily complex issues at play to provide comprehensive, nuanced insights toward converting the results of HPTN 052 into improved policies where HIV care is concerned. Using mathematical modeling to explore cost, potential efficacy, and other factors, this collection serves as a springboard for policy changes that could dramatically reduce the future spread of HIV, the virus that causes AIDS.

Current HIV treatment guidelines by the World Health Organization and other outlets call for ART to begin when CD4 cell counts reach at or under 350 cells per microliter. Typically, the number of CD4 cells — part of the infection-fighting immune system — decrease as the disease progresses. HPTN 052 found a benefit to starting treatment earlier, when CD4 counts are above 350 cells per microliter. Specifically, among heterosexual couples where one partner is infected and the other isn’t (so-called serodiscordant), earlier treatment reduced the rate at which HIV spread to the uninfected partner.

But moving toward earlier ART as a blanket policy isn’t so easy, and isn’t necessarily the next logical step. As Bärninghausen and colleagues write in their contribution to the PLoS collection, “HIV Treatment as Prevention: Issues in Economic Evaluation,” earlier ART can have negative consequences. If patients do not adhere to treatment, the disease can become resistant later on, making recovery less likely. The drugs have side effects that often require medical attention, adding to the physical burden earlier than may benefit the patient, even though it may halt viral spread. An individual’s economic productivity may also be curbed by ART therapy. And, the notion of giving drugs to someone infected with HIV as a way to prevent new infections, even when it might not benefit the person taking the medication, raises ethical concerns.

Yet the potential for treatment as prevention to stop the spread of HIV cannot be denied, as HPTN 052 made clear. In addition to a decrease in the number of HIV cases, earlier ART means patients won’t have to wait until their health begins to deteriorate before entering care. Economically, patients who start ART sooner rather than later may avoid that phase of deteriorating health and therefore keep working and active.

As the PLoS collection makes clear, the central question seems to be how best to implement the approach. In their article, “HIV Treatment as Prevention: Optimising the Impact of Expanded HIV Treatment Programmes,” Delva and colleagues provide a detailed analysis of how to prioritize the various sub-populations of those infected with HIV for treatment-as-prevention interventions. As they write, prioritizing patients according to CD4 cell count — initiating treatment at 350–500 cells/microliter — may not prevent new infections; patients with lower cell counts tend to have more highly transmissible disease. The cost of such a measure is significant because it would add 20% of the HIV-infected patients to care. Despite its questionable value, however, this approach may be the most acceptable because treatment need and access has historically been determined by CD4 count. The following figure, from Delva and colleagues, shows HIV transmission and mortality by CD4 count.

Prioritizing patients according to viral load might actually make more sense. Evidence from serodiscordant couples shows that infectiousness increases with increasing viral load (ART reduces viral load, which is why those on therapy are less contagious), though not drastically. There is also persuasive data showing that people with a high viral load progress rapidly to AIDS, pointing to a severe need for urgent treatment. But without more solid evidence on its clinical benefit (as opposed to its epidemiological benefit related to the surrounding uninfected population), stratifying by viral load is unlikely to gain traction.

Delva et al also evaluated the usefulness of prioritizing pregnant women, those with active tuberculosis disease, those in a serodiscordant long-term relationship, female sex workers, men who have sex with men, and people who inject drugs. Stable, serodiscordant relationships seem an obvious place to start because of its clear ability to reduce the spread of HIV. Yet data suggest that countries with the highest HIV rates tend to have the lowest rates of such couples, and finding those individuals may be difficult. Plus, prioritizing people in a serodiscordant relationship over those in a concordant relationship (both infected) may meet resistance. Theoretically, prioritizing female sex workers for early ART could have the greatest epidemiological benefit. The approach is also feasible, with previous intervention efforts having met with success. But treatment retention tends to be low among this subgroup, and prioritizing this group would be controversial.

Here is where mathematical modeling comes in (as was covered in the first post here on this collection). Several papers in the PLoS Medicine collection examine ART treatment as prevention using various models, an approach that enables a faster and cheaper analysis than would a clinical trial. According to Eaton et al, who compared 12 different models in their contribution to the collection, although the models vary in structure, complexity, and parameters, “all suggest that ART, at high levels of access and with high adherence, has the potential to substantially reduce new HIV infections.” The models suggest that an ambitious scale-up of earlier ART could have a dramatic short-term epidemiological impact. But when it comes to the long-term benefit, the results are more varied. In the following figure, (B) shows the reduction in HIV incidence by the year 2020 when access is expanded to 80% of the HIV-positive population with 85% retention of care.

If there is one constant note rung throughout this collection, it is that determining the best approach for HIV treatment as prevention is extremely difficult. But the papers also make clear that some change in policy is warranted, that expanding access to ART will, somehow, reduce the spread of HIV.

At least, that is the vision. Kumi Smith and colleagues add their analysis of ecological observations to the mix. With this type of study, observational data are used to shine a light on the link between exposure and outcome at the level of populations, rather than individuals. As the authors write, “[a]lthough we expect an impact of ART at the population level, the magnitude of the effect may not be as great as some hope…” The barriers to treatment as prevention may also diminish its potential, says David Wilson in his paper, “HIV Treatment as Prevention: Natural Experiments Highlight Limits of Antiretroviral Treatment as HIV Prevention.” Below, Wilson shows steps required in order to reduce onward transmission from someone infected with HIV:

The task at hand may be to find the middle ground between the potential benefit shown in many mathematical models and the difficult reality chronicled in natural experiments and ecological analyses. As The HIV Modelling Consortium Treatment as Prevention Editorial Writing Group (including many contributors to this collection) states in an introduction to the collection, “[t]he question of how to best use the tools that have been shown to reduce HIV transmission will likely dominate the field of HIV prevention for the foreseeable future.”

Category: HIV | Tagged , , , , , | 1 Comment

Mathematics and HIV

In anticipation of the 19th International AIDS Conference — being held in Washington, D.C., July 22-27 — Work in Progress is turning its attention to the upcoming collection of articles on HIV Treatment as Prevention published this week. I’ll have several posts here about the collection, looking at some of the concepts behind the work being presented, and then delving into specific articles in the collection. (See a Los Angeles Times article on the collection here.)

Woven throughout the collection is the use of mathematical modeling as a way to evaluate the impact of potential interventions that could diminish the spread of HIV/AIDS. Last year, the landmark HPTN-052 study showed that early antiretroviral therapy (ART) for people infected with HIV could reduce the risk of the sexual transmission of the virus to an uninfected partner by 96 percent. (The complete report of this breakthrough study, published in the New England Journal of Medicine and named “Breakthrough of the Year” by Science, is available online here.)

In the wake of that finding — 96 percent! — the World Health Organization and other agencies are trying to figure out the best way to implement ART as prevention (the concept of treatment as prevention will be the focus of the next blog post on this collection). And the best way to figure out that best way is through mathematical modeling.

Thirty-three million people were living with HIV in 2009, according to UNAIDS. Considering the scope of the current HIV epidemic, the mathematical modeling of ART treatment as prevention could spur the policy-making powers that be into actions that could make a dent.

So, onto the important questions: What is a mathematical model? How are they being used to study HIV treatment as prevention? What are the potential problems with these models in this arena?

What is a mathematical model?

Like any equation, a model is like a system represented by numbers and mathematical functions. By applying numbers to all the variables within a system — such as, the number of people infected or uninfected, the rate at which new infections may occur in the presence versus absence of treatment — it is possible to generate a numerical value that reflects the real-world value of whatever is being studied (in this case, the role of ART in preventing new HIV infections).

The usefulness of models depends on how well defined any particular factor is. Sexually transmitted diseases are particularly well-suited for modeling because the route of contact is clearly defined: it’s sex. In addition, there are so many questions at play when it comes to addressing the spread of HIV that modeling is really the only practical approach. The broad range of the epidemic also means that modeling is much more feasible than studying interventions in people over several years. And because the outcome of an HIV study can mean life or death for many, many people, you want to get those studies done as quickly as possible. Although the models being applied to studying ART treatment as prevention do have holes, the insights provided are much better than waiting five years for more study results. Plus, modeling is a lot less costly.

How do mathematical models work?

Mathematical models come in several varieties. They can be compartmental or distributional, where the former groups together all people with an infection and the latter describes a gradation of symptoms. They can be discrete or continuous, depending on whether you want to examine theoretical changes in a population as a smooth, continuous process or in chunks of discrete steps. Deterministic models are not subject to chance, whereas stochastic models incorporate chance into the equation.

Then there are more factors to consider. Do you want to study a population average or do you want to simulate how life is for an individual? Do you want a linear model or a nonlinear model? A linear model often has quite predictable results because the variables involved are linked tightly together: effectively treating one individual reduces the number of cases of a disease by exactly one (as would be the case with, say, breast cancer). HIV and other infectious diseases are modeled as nonlinear. Do you want an analytical or a numerical solution? Analytical solutions can show exactly how a given intervention will impact a community. But with complex problems like the spread of HIV, where sex, age, and sexual activity must all be factored into the equation, the solution must be numerical.

These are a few of the considerations that go into deciding on what kind of mathematical model to use. For more about these points, this article is a good starting point, as is this one.

For more about the application of mathematical models in addressing HIV, the HIV Modeling Consortium is a good resource. (Much of the work published in the PLoS Collection was done in collaboration with this consortium.)

What have mathematical models shown about ART treatment as prevention?

The PLoS collection on HIV Treatment as Prevention includes a comparison of 12 different models evaluating the impact of ART treatment on preventing further infections. That comparison found that the incidence rate would be reduced by 35% to 54% if ART were given to 80% of individuals with HIV treated after their CD4 cell count reached 350 cells/microliter. The graph below, from the introduction to the PLoS collection, shows the cumulative distribution of new infections generated by a single HIV-infected individual over the course of their life since being infected, in the absence of treatment (the red line denotes the start of treatment).

CD4 cells are white blood cells that fight infection. Their quantity decreases as HIV progresses. As Hallett et al write, current ART treatment tends to be initiated when CD4 cell counts are well below 200 cells per microliter, diminishing the potential for ART treatment to prevent new infections. Most transmissions would have occurred before that cell count is reached. According to mathematical models, the number of new transmissions from an infected person could drop dramatically if treatment is started closer when counts are closer to 350 microliters.

This graph, from the evaluation of 12 models by Eaton and colleagues that is part of the PLoS collection, shows the impact of treatment when ART is initiated at 350 CD4 cells per microliter, with 80% access and 85% adherence to treatment:

And this, from the same article, shows the proportion of HIV reduction by the year 2020, according to each of the 12 models included in the analysis:

As Timothy Hallett and others write in the introduction to the collection, “If the average number of new infections arising from an infected person in a susceptible population exceeds one before treatment could be feasibly initiated, then treatment could not eliminate the HIV epidemic.” The models examine the initiation of ART therapy at different CD4 cell count starting points, and then factor in other preventive measures, such as circumcision, or a decrease in the number of new sexual partners. Through those calculations, policymakers can obtain a view on what policies make the most sense for preventing the continued spread of HIV. The current World Health Organization guidance stipulates that HIV-infected people begin ART when CD4 cell counts reach less than 350 cells per microliter (though patients with advanced disease or with advanced tuberculosis should receive ART upon diagnosis, regardless of cell counts).

Problems with HIV treatment as prevention modeling

As Hallet et al write, the HIV testing rate used in the models evaluated in the PLoS collection is much higher than the 52% reported in the South African National HIV Prevalence, Incidence, Behavior, and Communication Survey (available as a PDF). And, the models assume the link between testing and ART uptake to be 100%, whereas that figure is probably closer to 33%. The dropout rate from treatment programs is assumed to be about 1.7% in “the most optimistic model simulations,” according to the article by Hallett and colleagues. The real-world dropout rate is probably closer to 10%.

Another huge hurdle is cost. Hallett and colleagues write, “After years of rapid growth, funding commitments and disbursements have stabilized or been reduced, and only a few countries in sub-Saharan Africa are currently able to achieve the high levels of treatment coverage for those eligible recommended by current international guidelines.” Programs with high costs in the short-term might be impossible to implement, regardless of how convincing a case is made by a mathematical model.

To that end, the PLoS collection includes an article by Meyer-Rath and Over outlining what economic considerations should guide discussions about programs that may be spurred by what mathematical models reveal. The cold reality is that “to a decision-maker, savings that are accrued in the future may be worth less than those made today.” In other words, it’s hard for many agencies to implement expensive programs because they will save money in the long-run. In that case, “…potential future payoffs may be less attractive, and investment in programmes for other, more immediate causes of mortality would be a rational, if not necessarily an inspiring or ethical, response.”

What next?

Hallett and colleagues note that a gradual expansion of access to treatment may be the best middle-ground solution. Then, policymakers can also look at which groups to focus on in order of priority. Pregnant women who are already under care may be easier to reach for ART initiation than other groups of people infected with HIV. Stable couples might adhere better to treatment than other groups. Knowing that part of the goal here is to prevent future infections, should “serodiscordant” couples (that is, one infected and the other not) be placed as the top priority?

In concluding their introduction, Hallett and colleagues note that mathematical modeling alone is not enough to guide decisions on HIV treatment as prevention. Epidemiology, economics, demography, statistics, and biology must also weigh heavily into all discussions. But in light of the HPTN 052 findings, mathematical models are providing crucial insights about the potential impact that earlier treatment could have on slowing the spread of HIV — or perhaps stopping it altogether.

Here is an interview with Timothy Hallett on the PLoS Blog, Speaking of Medicine, providing further insights about the newly published collection.

_____________________________________
All images courtesy of PLoS Medicine

Category: HIV | Tagged , , , , , , , , | 7 Comments

More on HIV and Poverty in the U.S.

Many years ago, some friends of mine in Israel dug a hole. I don’t remember now what they were searching for, but they thought there was something a ways down, so they dug. It took a while, and in the end they had a very deep hole. But, what they had been looking for turned out to not be there. So now they had a hole with nothing to show for it. Someone came up with a very surprising way to make use of the hole. Again, sorry, I don’t remember the details, but it was something wonderful. And the wisdom of the moment was: Why waste a good hole?

My last post was a brief survey of the landscape connecting HIV and poverty, with a selection of slides from a talk I was about to give at the American Association for the History of Medicine meeting. Having gathered together the information, I thought I’d post the remainder, because why waste a good slide.

You know how there are some topics which, when you come across them, you think: why aren’t we talking about this every day? How can the world just kind of keep going on when this problem exists? How is it possible that someone is spending more than $100 million dollars to keep The Scream in a private collection, or to ensure someone’s election into office, when there are children being born into situations that put them at high risk for diseases that they’ll spend their entire lives coping with, or from which they will die way too young? It’s not that I think the people spending that money should be directing it elsewhere; it’s that we live in a world where economy trumps humanity at every turn. I know, I know, this is a blog about science, not social justice, but sometimes a topic arises where the two can’t be kept separate.

There are a zillion issues like this, too many to name, or almost think about. Admittedly, delving into the many health disparities connected with poverty isn’t something I do every day. But having gone there, I’m finding it hard to leave.

Anyway. Onto the slides.

In the last post, I included some stills from this moving portrait of how HIV has shifted over the past two decades. Here are some more statistics about that:

The AIDSVu project, conducted by Emory University, found that nearly all U.S. counties with high rates of HIV and infection are located in the south. To be more exact, of the 175 counties in the top 20% for HIV and poverty, all but six are in the south.

According to the 2010 CDC report, Communities in Crisis, heterosexuals living below the poverty line in U.S. cities are five times as likely as the general population to be HIV-positive, regardless of race of ethnicity. Among people living in the same neighborhood but above the poverty line, the likelihood of being HIV positive was 2.5 times higher than among the general U.S. population.

Here is another way to consider the overlap of poverty and race/ethnicity:

So, the CDC is saying that the main connection between HIV and race/ethnicity is through the conduit of poverty. When you look at the entire U.S., HIV rates are staggeringly higher among blacks. But when you zero in on poverty, the racial/ethnic disparities almost disappear. Similarly, sexual orientation becomes less of an issue. In that CDC study, which was based on 9,000 people in 23 cities, 2.1% of heterosexuals living in high-poverty urban areas were HIV positive. Note that the definition of an epidemic is when the rate of a disease in a given location exceeds 1%.

The last post also mentioned that HIV is one of many diseases of poverty. Of course this is looking only at the U.S.—when you extend the view to the entire world, the connection becomes only more harrowing, as shown very starkly in this WHO report, Disease of Poverty and the 10-90 gap. (PDF) (10-90, as in: 10% of the world’s population get 90% of the healthcare.)

In the U.S., diabetes is another disease connected with poverty. Here is a slide that was included in the last post showing the prevalence of diabetes in the U.S.:

The map above covers all regions of the country where income spans from $19K/year to $112K/year.

Now here is a version of the map filtered to include areas where the average income ranges from $19K/year to $34K/year:

The two maps reveal the areas where lower-income plus diabetes incidence are most concentrated.

A major issue inside this territory is that of education. Educational disparities are also linked to disease. Having not really done much research to better understand, in a clear and factual way, why poverty and education and disease seem to be linked together, I’m not going to assume any explanation. But here are some small glimpses that show the link exists.

I hope you can see this slide clearly enough. It shows the potential number of deaths from cancer that could have been prevented by eliminating educational and/or racial disparities among people ages 25–64 in 2007:

The light blue in the charts above, which you’ll note encompasses more than half the pie for African American men, shows the deaths that supposedly could have been avoided.

Among the 10 states with the lowest high school graduation rates (less than 65%), seven are in the southern U.S. Whether that bears any direct relevance to HIV rates in those states, I don’t know.

Here’s some more interesting information, now focused on sex education. There are a lot of states that don’t mandate sex education or HIV education.

For a better view of that, a PDF is available here, with a lot more information about where American teenagers get information about sex. Several states in the U.S. allow abstinence-only sex education.

Here is more about the link between education and health, from the National Poverty Center (PDF). 

Finally, some more information about healthcare among people living with HIV.

The CDC recommends HIV screening as part of routine medical care. Yet in many states, Medicaid does not cover routine HIV screening. According to the CDC, “States in the South were least likely to cover routine HIV screening (4 of 16).” This map shows where such screening is (dark blue) and isn’t (light blue) covered:

Finally, a look at poverty rates across the country. According to the U.S. Census Bureau, rates are rising faster in the southern U.S. than in other regions. From 2009 to 2010, there was a 1.2% rise in poverty in the South, about double the rise seen in the Northeast, Midwest, or West. The share of the U.S. population earning below half of the federal poverty line has risen to 6.7%, a record high.

Category: Healthcare disparities, HIV | Tagged , , , , , , , , , | 2 Comments

HIV and Poverty: A Slide Show

This Saturday, April 28th, I’ll be joining documentary filmmaker Lisa Biagiotti and Stephen Inrig, a professor at UT Southwestern Medical Center, on a panel session at the annual meeting of the American Association for the History of Medicine. Our talk will focus on HIV in the Southern U.S., with my portion focusing in particular on the connection between HIV and poverty in the region.

In creating some slides to show during the talk, I was struck yet again by the starkness of this connection, and of the deeply engrained link between socioeconomic status and health. Because I know many people feel likewise alarmed and simultaneously glad to be aware and reminded about these connections, here is an extract of the slide set, with some explanations here and there.

The U.S. Census Bureau includes the following states/regions in its definition of the southern U.S.: Washington D.C., plus Alaska, Arkansas, Delaware, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, South Carolina, Texas, Tennessee, Virginia, West Virginia.

Here’s a look at the distribution of wealth in the U.S. According to the U.S. Census Bureau, of the 12 states with >30% of its population living in poverty, 12 are in the South:

Viewed from another angle, the five states with the lowest personal income per capita in 2007 (which I realize is several years ago now) are all in the southern U.S:

When HIV was first found in the United States in 1981, it was concentrated mainly in coastal cities, and did not turn up in any particular population in terms of race or economic status:

This moving image created by Scientific American, in conjunction with an article I wrote for that magazine about HIV in the South (Poor Man’s Burden, with much credit to editor Christine Gorman), chronicles the shift in geography that HIV has undergone in the United States since 1981. Here is the 2009 view:

Scroll back and forth between those two images to see how the orange circles have changed location.

Over time, HIV has come to be concentrated in areas stricken by poverty. To be sure, there are other factors that have contributed to the rise in HIV in the southern U.S.—social stigmas, for example—but economics is a major issue, and one that is tightly linked to race. Here are some statistics about HIV in the southern U.S.:

Here is a particularly stark view, showing HIV levels in several countries:

And let’s remember, the U.S. is the country that spends the most on healthcare in the entire world.

Here’s one more look at the link between HIV and income levels. Note the quote from the CDC’s report, “Communities in Crisis,” stating the inverse link between poverty and annual household income.

The connection with race is also clear, as in this chart and the statistics noted below:

In case it’s hard to read, here are some statistics:
• More than half of all poor and black households are in the south
• Blacks are 13% of the U.S. population and 51% of people living with HIV/AIDS
• The estimated lifetime risk of HIV is:
- 1 in 16 for black males
- 1 in 30 for black females
- 1 in 104 for white males
- 1 in 588 for white females

Of course, HIV isn’t the only illness linked to poverty. Diseases of poverty also include heart disease, diabetes, obesity, and some types of cancer (as former director of the NCI Samuel Broder once said, “poverty is a carcinogen.”). Here’s a look at the distribution of diabetes in the United States:

There are numerous reasons why poverty is linked to disease, and HIV in particular. Poverty is associated with inadequate education, limited healthcare, and inadequate career opportunities. The limit on career opportunities can prevent people from obtaining the independence needed to resist risky behavior. Poverty is also associated with higher rates of incarceration, which has a dramatic impact on the surrounding community. The quote in the slide below says so much:

Regarding the access to healthcare, poverty restricts it. In many states, Medicaid eligibility sets federal poverty level limits that leave many people ineligible, but yet unable to afford private insurance. Here is a look at the percentages of uninsured people under 65 and children under 18 in the United States in 2008:

Another problem with Medicaid is that routine HIV screening is not covered in many areas. The CDC reports that only 4 of 16 states in the South routinely cover HIV screening, even though this approach is recommended, especially considering that among the 1.2 million people currently living with HIV/AIDS, about 1 in 5 don’t know they are infected. Here is a look at where routine HIV screening is and isn’t covered by Medicaid (dark blue = yes, light blue = no):

In several states, Medicaid has limits on the number of prescription drugs allowed (often not enough to cover all drugs needed for proper HIV care) and other healthcare factors.

Access to primary care physicians is also a problem, though not as dramatic as in rural areas. Here’s a look at “health professional shortage areas,” specifically primary care, according to HRSA:

Poverty rates in the south are rising faster than in other regions of the country. The slides above are a sampling from my talk, which in itself offers only a glimpse at the many issues involved (I’m not an expert, it should be said). But hopefully they provide some insights into a complicated and very serious problem.

Category: HIV | Tagged , , , , | 8 Comments

Avastin for Lung Cancer

One of the most interesting aspects of a newly published study of bevacizumab (Avastin) for older patients with advanced non-small cell lung cancer (NSCLC) is the abstract’s conclusion:

“Adding bevacizumab to carboplatin and paclitaxel chemotherapy was not associated with better survival among Medicare patients with advanced NSCLC.”

It’s refreshing to see no extra flourishes added, no unnecessary words trying to imply that somehow the drug might be of some benefit to some patients maybe. Of course better treatments are needed for older people suffering from advanced lung cancer, but there’s no use forcing a combination if it doesn’t work.

The study, conducted by several physicians at the Dana-Farbar Cancer Institute, was actually a retrospective examination. A total of 4,168 patients with NSCLC, all age 65 years or older and all Medicare beneficiaries, all diagnosed between 2002 and 2007, were split into three groups, as follows:

Group 1: Diagnosed in 2006-2007; initial chemotherapy with bevacizumab-carboplatin-paclicataxel

Group 2: Diagnosed 2006-2007; initial chemotherapy with carboplatin-paclitaxel (so same as above minus bevacizumab)

Group 3: Diagnosed 2002-2005; initial chemotherapy with carboplatin-paclitaxel

Overall survival measured from the first date of chemotherapy treatment until death, or the “censoring date” of December 31, 2009, served as the primary outcome of the study.

According to the report, just published in JAMA, the median survival estimates were:

Group 1: 9.7 months; interquartile range, 4.4–18.6
Group 2: 8.9 months; IQR, 3.5–19.3
Group 3: 8.0 months; IQR, 3.7–17.2

One-year survival probabilities were:

Group 1: 39.6% (95% confidence interval, 34.6%–45.4%)
Group 2: 40.1% (95% CI, 37.4% –43.0%)
Group 3: 35.6% (95% CI, 33.8%–37.5%)

Importantly, the authors state in the abstract: “Subgroup and sensitivity analyses for key variables did not change these findings.”

Bevacizumab is already approved for advanced NSCLC, where it is given in combination with carboplatin and paclitaxel for patients who have not received chemotherapy for advanced disease already. (The drug is also approved for the treatment of metastatic colorectal cancer, metastatic kidney cancer, and glioblastoma.) This new report seems to suggest that the triple combination for NSCLC should probably be confined to patients under 65 years of age. (NOTE: I am NOT a doctor and am not qualified to offer any opinions on cancer treatments.)

Numerous clinical trials of bevacizumab for cancer are ongoing. Some are sponsored by competitors seeking to compare their drug to a bevacizumab-containing regimen, and others are focused on combining bevacizumab with other agents such as AZD2171, temozolomide, Revlimid, and many others.

And for anyone seeking additional context about bevacizumab in the treatment of cancer, here are some of my previous posts. Looking at this list, you might begin to question the sanity of a writer so interested in a single drug. But tracing the history of this drug provides powerful insights about clinical trials, the drug development process, FDA review of new drugs, weighing benefits against side effect risks, and how to consider cost in healthcare decisions. When a single thread tells this much of a story, it’s worth following.

Avastin for Ovarian Cancer
Let the Death Panel Accusations Fly! (On the withdrawal of FDA approval of Avastin for breast cancer)
Taking a Stab at Cost-Effectiveness
The Avastin Saga

Category: cancer, Clinical Trials | Tagged , , , , , , | 5 Comments

New Study of HIV Hot Spots

HIV infection rates of black women in certain parts of the United States is five times higher than the overall rate of infection among black women, according to a newly published study by the HIV Prevention Trials Network (HPTN).

The study, HPTN 064, looked at HIV rates in six geographic “hot spots”; that is, regions of the U.S. known to have elevated rates of HIV and poverty. Just how elevated are those HIV rates, the study aimed to address.

The hot spots included in the study were Atlanta, GA; Raleigh-Durham, NC; Washington, D.C.; Baltimore, MD; Newark, NJ; and New York, NY.

According to a report of the study in Infection Control Today, about a quarter of new HIV infections in the U.S. occur in women. Of these women, 66% are black—a figure that stands in stark contrast to the fact that black women account for 14% of the U.S. female population.

HPTN 064 (also called “The Women’s HIV SeroIncidence Study, which somehow qualifies for the questionable acronym “ISIS”) enrolled women ages 18 to 44 without a prior positive HIV test. The stated purpose of the study was “to estimate the overall HIV-1 incidence rate in women at risk for HIV acquisition in the US and to evaluate the feasibility of enrolling and following a cohort of these women.”

Among the 2,099 women enrolled in the study, 88% of whom were black, the HIV incidence was 0.24%. (As a measure of comparison, here is a chart showing HIV/AIDS rates in sub-Saharan Africa, though note that the rates show HIV/AIDS, not non-AIDS HIV) Enrollees in HPTN 064 were asked about their mental health, sexual behavior, history of sexually transmitted infections, domestic violence, social support, financial insecurity, and health care utilization.

Thirty-two women were found to have HIV infection at the time that they enrolled in the study; they had been previously unaware of their HIV status.

The findings of HPTN 064 have not been reported in the New York Times, the Washington Post, or the LA Times, even as an online blog item.

Among the study sites were Johns Hopkins HPTN Network Laboratory, Bronx-Lebanon Hospital Center, Harlem Prevention Center, New Jersey Medical School, Wake County Health and Human Services, University of North Carolina AIDS Clinical Trials Unit, The Ponce de Leon Center, and the Hope Clinic of the Emory Vaccine Center. The study was funded by the National Institute of Allergy and Infectious Diseases, part of the National Institutes of Health.

For a more graphic view of HIV/AIDS across the United States, the HIV/AIDS Atlas shows rates by county. (You have to do a quick, free registration to use the atlas.) The view is startling. Did you know that most of eastern Massachusetts has HIV rates of .174%–.309%? This is in the upper reaches of prevalence. You can also see the ultra-concentrated pockets of elevated rates around Kansas City, MO; Jackson, MI; and Little Rock, AR, among other places. In each of these areas, HIV rates are in the .174% to .309% range, while the immediately surrounding areas have much lower rates. See the bar graph on the site for a make-no-mistake view of the racial divide at play here.

________________________________________
Image (showing high rates of HIV in certain areas of New Jersey) from the HIV/AIDS Atlas

Category: HIV | Tagged , , , , , , , , | 1 Comment