Physicians for Clinical Responsibility

From ca. 2010-
The Bioethics of Avastin and Retinal Pharmacotherapeutics
Practicing medicine based on best available evidence seems obvious, but this has not always been done. Because medical decision making has in some cases been based on tradition or parochial convention, the recent trend has been to try to base decisions on objective evidence. This should mean all evidence. Real “Evidence Based Medicine” means that, as physicians, we are professionally charged and morally bound to intellectually consider ALL evidence; prospective, retrospective, local, and global, and evaluate it ALL for its strengths and weaknesses, and then do our best to arrive at a composite best decision for each individual patient.

What we have instead come to is “Evidence Based Marketing”. Because industry, and particularly the pharmaceutical industry, now controls funding of much of medical research, continuing education, and academic administration, they can exert heavy influence on just what “evidence” is promoted in medical decision making. Either through financial influence, and/or naïveté, some academicians and community doctors are willingly complicit in the process, and the honest efforts of the many who are not complicit is tainted in the process.

This has reached a crisis level in the design and promotion of numerous “Phase IV” trials offered by innumerable practices with little experience and no motivation beyond reimbursement and marketing themselves as “study centers”. The financial benefits of such practice is nowhere to be found in the patient’s informed consent, but to those who know the true nature of such efforts, it is that financial impact that has given rise to the vulgar title of “pharmawhore” to such doctors. This moniker isn’t entirely accurate. The doctor has not become the drug company’s “whore”, but rather its “pimp”. Just as a real pimp often sells young girls into prostitution with promises of glamorous careers in modeling or acting, these doctors literally sell their patients to drug company marketing interests under the illusion that they are participating, and may benefit from, some noble new research into new treatments. The financial relationship between the pimp and the “customer” is never fully disclosed to the patient, and they are never aware that the primary purpose of the “study” is to find new markets for insanely expensive products. It is the emergence of Avastin as not only a wildly effective treatment of various retinal conditions, but also its extreme cost-effectiveness, that has shattered the carefully cultivated paradigm that retinal pharmacotherapeutics can legitimately be obscenely expensive. Avastin threatens the prospect that drug companies can literally drain government and third party payer coffers, and the marketing of lucrative “Phase IV” studies to mercenaries is just one tactic in the pharmaceutical industry’s onslaught of medical care.

This is both shameful and unnecessary. The pharmaceutical industry may not be the miracle workers they make themselves out to be, but they do create many valuable therapies, and even without resorting to mendacity and even fraud, they could be a successful and well-respected industry. Instead of honoring their public responsibility, however, they deny it by preying on the very society that gives them their raison d’être.

Let’s make an analogy. The pharmaceutical industry has given us Prilosec, a very useful drug for gastric hyperacidity. When its patent was due to expire, the same company came out with Nexium, which
is Prilosec. The only difference is that Prilosec is the racemose mixture and Nexium is the more active isomer. On a dosage adjusted per weight basis, they are the same thing. (The other difference of course is the exorbitantly higher price of Nexium). The company benefits greatly through deceptive marketing that conceals this basic truth.

Or consider Lucentis, the ultimate “me too” drug, in that the original “me” version was never even promoted for an obvious use. The maker of Avastin either did know, or had the responsibility of knowing, that it would be expected to be safe and effective for macular degeneration, but they elected not to even test it. They did of course know that they had already priced it for cancer treatment in doses of several hundred milligrams every 2 weeks at a price point of around $50,000.00 per case per year. They also knew that eye treatment would require only 1 mg. or less and so there would be little revenue stream for this application. Although there was already extensive safety data from the cancer data in systemic doses hundreds of times higher, they elected to not even test Avastin, suggested that it was unsafe, obfuscated its potential claiming it was “too big to work in the eye”, split the molecule that they had every reason to think
would work, do some (as it turns out unnecessary) modifications and bring it out with a new name and a price of $2000.00 per 0.5 mg. vs. $5.00 per mg. for the parent drug. I say unnecessary because independent testing showed Avastin to be so obviously effective from the very first patient, and so obviously reproducible with hundreds of thousands of other patients, that it has become the clear standard of care around the world for both effectiveness and cost effectiveness. Although the company has refused any sponsored research comparing the two forms of this same drug, it now seems apparent from hundreds of thousands of off-label uses that the parent drug works safely and may be stronger and longer acting, as we would expect since the derivative product clears the eye too quickly. This company touts as its slogan that they are “Bringing our vision of science to the science of vision”, but it would be possible to argue that what they are really best at is “Bringing the science of marketing to the marketing of science”.

Now for the analogy. The Chrysler Corporation essentially invented the minivan back in the 1980’s. They had a great product, which they marketed well and from which they profited immensely. It was appropriate, legal, and to be expected that other carmakers followed suit, and a high quality, affordable market was born. What if Chrysler had the same anticompetitive protections as the drug companies? They certainly also spend millions on development, and they have as much or more external pressure as the pharmaceutical industry and the same claim of their importance to society. After all, who cares if society is well medicated if everyone is starving for lack of transportation to work? What if they could have patented the minivan
concept? Then they would have a 17 year monopoly, at the end of which they could alter it slightly, rename it, and essentially renew the monopoly. Or, what if they could say, “Well we’ve got this great car for the commercial use, but it would never work for individuals, so we’ll change the color and increase the price 50-fold for private individuals only”?

Well, that is exactly what is happening. The very industry that could be leading the charge toward excellent
and affordable health in the USA is in fact the single biggest force driving the system to bankruptcy. They are literally looting the system as fast as they can. Am I exaggerating? Hardly. Let’s take the Lucentis example again. Aside from the question of whether it ever needed to exist at all (which it now clearly did not), just how egregious is the pricing? If cancer patients were charged for Avastin what eye patients are charged for Lucentis, each dose of Avastin would cost around $500,000.00. That’s each dose, given every 2 weeks, just for this one drug. How is this not surreally greedy and irresponsible? As another metric, consider that in recent surveys as many as 79% of doctors, out of responsibility to do the best for their patients, choose Avastin over Lucentis. In spite of this, Lucentis cost close to $1,000,000,000.00 (that’s 1 BILLION dollars) in 2007 alone. That is just one drug and it is just for about 20% of only one of the many conditions that can be treated with this. This drug is injected and so comes out of Medicare Part B. One billion dollars is over 20% of the entire 4.6 billion dollar CMS budget for all of eye care in the US. This budget is under extreme and increasing pressure, and is currently scheduled for a 40% decrease from 2008 to 2012 before considering the draconian cost of Lucentis. It is clear that the same public benefit can be safely achieved from perhaps $10,000,000 worth of Avastin. During this same time, the maker of Lucentis (and Avastin) is seeking through misleading direct-to-consumer marketing and outright coercion of compounding pharmacies to increase the market share of Lucentis drastically. If this agent was used according to labeled use for AMD, this one indication alone would cost about 5 billion dollars, more than the entire budget for all of eye care. They simultaneously claim that we must NOT use Avastin since it is “off label” and claim that we can use Lucentis in lower “off label” dosing. The hypocrisy in interpreting available data is breathtaking.

This Lucentis story is only for AMD, only for new yearly cases of AMD, and Lucentis is only one of many exorbitantly priced drugs in eye care. Another company makes a sustained release implant containing a 10 dollar amount of steroid for which they charge $18,250.00. Furthermore, Lucentis is just one drug in just one disease, in just one fragment of one specialty. Unbelievably, the documented examples in psychiatry, cardiology, gastroenterology, and emergency medicine are even worse. We are literally looking at the wholesale looting of the US health care system by the most anticompetitive, government protected, and government infiltrating industry in American history.

Redefining “Medical Evidence” for marketing purpose-

How do we get from “Evidence Based Medicine” to “Marketing Based Evidence”? How is the thought process transmogrified? In a word, subtly. Through careful marketing, industry has subtly changed the rules about what research to follow to an audience only too happy to be told what to do and/or paid to go along. Frankly, it is easier to be told what to do than deciding for yourself through an independent reading of the data. You just need to trust the source, which for some is the drug company and for others is the “Key Opinion Leader” being paid by the drug company. Unfortunately for industry and its minions, the practice of picking and choosing data breaks down when one overplays one’s hand. Although this process has been insidious, the Lucentis story demonstrates such an amalgam of greed, licentiousness, and illogic that we must now address that which has become untenable.

In the retina world, Evidence Based Marketing succeeds with the help of certain “myths” that have been promoted since the marketing of Visudyne. They are promoted by many in the drug industry and sponsored research community. These issues are exactly those used by corporate advocates and parroted by a surprising number of doctors fearful of bucking the corporate sponsored party line.

A governing theme is the growing “myth” in recent years that
only prospective randomized clinical trial data is real data. This, along with the more insidious myth that corporate sponsored randomized controlled trials (CSRCTs) are just as legitimate and unbiased as independent RCTs (IRCTs), seems to imply that retrospective data is useless. In other words, the myth implies that we can never learn anything useful or normative from past experience. If one can completely and categorically dismiss past experience and, through seizing financial control of the R&D pipeline, govern prospective data management, then one can define the “evidence” that governs future care according to one’s marketing plan. That is exactly what has been happening, and either out of fear of industry power or gratitude for millions of dollars of “Direct to Consumer” marketing, the mainstream media has been only too pleased to go along.

Contrary to these myths, there are some basic truths about RCTs and retrospective or experiential data that we need to consider. Frankly, it is easier to follow the crowd. It is easier to assume that any data from an RCT is unimpeachable and that any other data is useless, but we cannot afford to be that lazy. We are professional caregivers and with that responsibility, we need to consider and critically evaluate all data and recommend to our patients treatments based on all evidence. I am not suggesting that we consider retrospective data in preference to prospective data, only that we are responsible for critically analyzing ALL data according to its strengths and weaknesses. That is what Evidence Based Medicine requires and what Evidence Based Marketing seek to prevent.

We therefore need to take a hard look at the soft underbelly of corporate sponsored RCT data:

Myth #1- All RCT data is equal and inherently unchallengeable.

Truth #1- RCTs are only as good as the premises upon which they are based and the questions that they ask. If insightful, honest, non-leading questions are designed, it is more likely that useful, normative data will be gained. If uninsightful, or worse, self-serving questions or premises are employed, then it is easy to manipulate RCT data and alter the apparent outcome. The data, although statistically legitimate, can be made to only answer the questions and suggest the outcomes that you want it to. One example of course is found in of the
absence of questions about histopathology in the original PDT studies and the decision to redefine “visual success”. As another example, the experiential data on Avastin is so strong that it is virtually impossible to imagine that the brilliant, creative, and careful people at Genentech could have missed its utility unless they intended to. Although this is only speculation, it makes the parallel Lucentis development look like a contrived effort to circumvent the pricing of 1 mg. of Avastin for ophthalmic use, when the price of hundreds or thousands of milligrams had already been fixed for oncologic use. It is now abundantly clear that the premise upon which Lucentis was developed, that Avastin would not work in the eye because the molecule was too large, is wrong. Whether intended or not, this false premise abrogates the need for Lucentis to exist at all. Ironically, the only value of the Lucentis trials is to indirectly suggest that its analog Avastin is probably also safe and effective. The unintended implication of the Lucentis trials, and particularly the PIER trial9 which shows poor utility of Lucentis in dosage frequencies that seem to work with Avastin, is that Lucentis may indeed be inferior to Avastin regardless of price.


Myth #2- Any treatment can be evaluated in an RCT.

Truth #2- Only simple standardized and relatively non-skill-intensive treatments lend themselves easily to an RCT. Treatments that cannot be easily standardized, particularly skill intensive procedures, are difficult to study in an RCT. There are many treatments that have become standard care without the benefit of a prospective trial, such as insulin, penicillin, and now Avastin, where experiential data decisively removes doubt as to safety and utility. There are other examples of therapies, especially surgeries, for which trials span generations of surgical refinement and end up not recommending techniques that had subsequently became unequivocal standards of care because of refinements not included in the study design. The evolution of diabetic vitrectomy in spite of marginal DRVS data stands as one example.

Myth#3- Pathologic conditions are nonvariable and therefore the spectra of severity and pathologic character do not affect the quality of prospective studies.

Truth #3- Pathologic conditions are highly variable, and RCTs are actually only useful in studying very specific conditions within a disease spectrum. Beyond that, we as physicians have a longstanding tradition and a very legitimate responsibility to use our insight, best judgment,
and past experience to apply what we have learned from data available to us, whether from an RCT or elsewhere, as we treat patients suffering from a wide variety of disease types. In point of fact, RCTs can only provide basic direction , and individual and reported past experience is far more valuable and far more appropriate data to use in addressing the wide variability of disease that we encounter, variability that is not specifically addressed in the RCT.

Myth #4- RCT data is valid and retrospective data simply is not. We should therefore only consider prospective data (uncritically) and ignore retrospective data (categorically). This is a myth that drug companies have fought hard to promote in regard to CSRCTs in recent years, and incredibly, it is the most common argument I hear from those who choose Lucentis over Avastin.

Truth #4- RCT data when critically analyzed for its strengths
and weaknesses is valuable, as is retrospective data. We simply need to critically analyze and use data for what it is, not giving it greater weight than is deserved (as by those who uncritically accept RCT data and only RCT data), and not ignoring any data that is carefully collected. The whole subdiscipline of Bayesian analysis considers the validity of outcomes based on the application of conditions retrospectively. Our most legitimate use of RCT data is to use it within its limitations and to use our best judgment and experience to make treatment decisions for our patients. A compelling recent example of course is again the situation with Lucentis and Avastin. The limited RCT data that is available says that Lucentis is safe and useful in the treatment of exudative AMD. The drug company, its minions, and those unwilling to accept the responsibility of practicing true evidence-based medicine would seem to like us all to leave it at that. The problem of course is that we know that Avastin exists and that it is in fact the “parent” analog of Lucentis and therefore might be expected to act similarly. We have individual and reported experiential data that says not only that it works but that it works astonishingly well with no sign of unexpected adverse effects. We have the further burden of knowing that it works so well as to raise the possibility that the antibody’s developer should have been able to know this before pronouncing it useless, legitimizing the analog’s development that in retrospect seems like little more than a strategy to circumvent pricing regulations. We have the unsubstantiated but likely possibility that the smaller molecular weight size of Lucentis may in fact be a liability, causing it to have too short a duration of effect compared to Avastin, possibly allowing a disease to smolder and that this may promote fibrosis and less durable treatment effect. Finally, we have the knowledge that many patients have significant financial exposure with Lucentis, when all available clinical data suggest that they can have the same or better effect with Avastin at a miniscule fraction of the cost and with, in our best judgment, the same degree of safety. Theoretically, Avastin might even be safer, since the smaller Lucentis molecule has more rapid and higher peak clearing into the systemic circulation. We are responsible for ALL of this data, and morally we do not have the luxury of letting those with a financial interest do our thinking for us. Reasonable people may conclude that the best evidence-based decision is to use Avastin over Lucentis, unless comparative studies, prospective and/or retrospective, suggest otherwise.

Ironically, Genentech may be the strongest believer of all in the retrospective Avastin data. It is only on the basis of this data that they could reasonably refuse to study Avastin head to head with Lucentis, believing (as seems reasonable) that Avastin would be shown to be at least as good and possibly superior to Lucentis at about 1/100th the cost per case. This is of course only speculation.

Myth #5- Independent RCTs (IRCTs) and Corporate Sponsored RCTs (CSRCTs) are identically nonbiased and provide identically unimpeachable data.

Truth #5- RCTs in general are limited by the quality of the questions that they ask and the elegance of the studies used to answer them. Consequently, no RCT is above critical analysis. Furthermore there is growing evidence in multiple medical disciplines that CSRCTs are significantly prone to bias and should be subjected to even more diligent scrutiny, especially considering the financial pressure to produce profitable products,. This was recently reported in regard to psychiatric drug testing 12.We have already discussed the Visudyne example, where the pragmatic definition of visual success and neglect of now obvious macular toxicity were beneficial to the acceptance of a marginal therapy. To this, we should add the question of whether Lucentis ever needed to be developed at all.


A New Standard of Care

With the emergence of nonselective VEGF inhibition, the standard of care in the treatment of wet AMD and other exudative maculopathies has changed abruptly and decisively. Because of the volume of retrospective data and the ubiquity of experience throughout the retinal world, Avastin has become the de facto standard against which all other treatments, including Lucentis, must be compared. Why is that? Because the de facto Avastin standard was set first. To put it colloquially, by the time Lucentis came along, the Avastin cow was already out of the barn. There are many precedents for rapid adoption of de facto standards of care. These include aspirin, insulin, penicillin, and many surgical procedures, to name just a few. Although its own maker appears to have made a conscious effort to ignore or even suppress its use in the eye, Avastin’s impact on various maculopathies is equal on an historic scale to these other examples. That it is the standard of care is well established in the ASRS 2007 Trend survey, published in November of that year.

Looking back on the various recent treatment developments for AMD, one of the things we have learned is that rapid and complete inactivation of the neovascular sequence seems to be extremely important for long term visual success. Without this, the neovascular and/or fibrotic processes continue, resulting in worsening vision and lower success rates after conversion to more effective therapy. We have learned that PDT actually
causes damage to the surrounding “normal” RPE and choriocapillaris on a scale comparable in some patients to thermal macular ablation, at least with infrared thermal laser 1-3. To this, we must add the risks of cataract and glaucoma associated with intravitreal triamcinolone, since “unprotected” PDT is even more clearly risky and inefficacious.

Like PDT, Macugen use is associated with continued visual loss in most patients and expansion of vascularity and fibrosis within the lesion8. The speculation is that this is due to the inability of VEGF165 inhibition to completely stop the neovascular process. Early data on studies of Lucentis9 suggest that decreasing the frequency of injection from 4 weeks to three months is associated with progression of the neovascular process. Perhaps because of its small molecular size and rapid clearing from the eye, its fairly brief duration of effect may allow progression of the neovascular/fibrotic process in the same way that Macugen’s selectivity mitigates its effect.

The experiential data on Avastin suggest that it causes an extremely rapid and complete cessation of neovascular activity. Furthermore, this effect seems to last at least two months in most patients. Although this needs to be quantified, we have seen noticeably less submacular fibrosis on exam and OCT than has been typical in Macugen or PDT treated patients. Histologic, Dynamic ICG, and ERG studies, both reported and in our practice, show none of the anatomic or functional disruption that is now well documented with PDT. We have also seen none of the choriocapillaris devastation, central scotomata, or florid vascular regrowth after Avastin treatment that is common after PDT.

All of this supports nonselective VEGF inhibition as the new standard of primary care and of Avastin as offering the best combination of safety, efficacy,
and cost-effectiveness. Initially, some payers required that patients fail with the older, and now clearly inferior “conventional therapies” before they would cover Avastin, but this is becoming rare as more and more have recognized the Avastin standard. We also see suggestions to continue using Macugen as maintenance after nonselective anti VEGF “induction”. This seems senseless, and appears to be a contrivance by Macugen advocates to find some use for a drug that is more likely to become an historical anecdote.

A Crossroads for Research-

In spite of this revolutionary change in our understanding of the neovascular process, and in spite of the growing acceptance of Avastin as the de facto standard of care for exudative maculopathies, many continue to describe PDT and Macugen as “standard” or “conventional” therapies for AMD. Furthermore, some researchers have so far continued to recruit patients for company-sponsored Phase IV studies that place PDT and Macugen as primary therapies or anchor therapies in combination protocols. Neither of these practices has any logical basis in the era of nonselective anti-VEGF therapy. These studies are based on a standard under which continued visual loss is acceptable, a standard which continues to be promoted by makers of inferior drugs, including providing material compensation to doctors who assist in these efforts. This practice has no justification other than to preserve revenue stream at patient expense and to benefit doctors through compensation and illegitimate research “prestige”. We simply cannot ethically continue to recruit patients into studies that we know will cause tissue destruction or indolent continuation of the disease process or where we expect vision loss to continue when we know that those same patients can receive safe, cost-effective treatment that causes no anatomic or functional harm and from which we can expect a higher rate of predictable and rapid improvement in vision.

We are at a crossroads that demands a complete re-evaluation of our investigational goals. Nonselective VEGF inhibition is not a cure for neovascular and exudative maculopathies, but it is vastly superior to any of the older treatments still in use. This is not to say that those older therapies cannot have an adjunctive role, although that may not be likely especially for the more invasive and expensive ones. What it does mean is that the visual and anatomic effect of non-selective VEGF inhibition is no longer an unknown. The early prospective data on Lucentis is definitive, and the retrospective data on Avastin is so strong that an RCT to evaluate its effect compared to natural history or the effect of PDT or Macugen would not be ethical. There are suggestions that Avastin may be superior to Lucentis as to its duration of effect, and of course it is obviously and vastly superior in its
cost-effectiveness. The relative utility of Avastin vs. Lucentis is still quantitatively unknown however, and since both have clearly similar safety and efficacy profiles, it is scientifically reasonable and ethical to compare them in a prospective trial. If nothing else, the vast positive experience with Avastin far exceeds the pilot data on Lucentis patients that the company determined was adequate to mandate further study.

Given the overwhelming experience with Avastin, even the head to head study seems destined only to confirm what we already know. Even if it is completed, it is best to do it independently. Genentech has been quoted in the national press on three different continents as saying that they will never support a comparative trial with Avastin. Given the recent concerns over bias in company sponsored RCTs, an independently funded and administered RCT would have more legitimacy, and Genentech’s noninvolvement may actually be an advantage.

In spite of all this, research as an extension of the marketing effort continues. With the imbroglio created by Genentech’s recent efforts to force Lucentis use by restricting Avastin availability, the ASRS leadership has finally and laudably taken a stand against such behavior. At the same time though, the 2007 ASRS “academic” program can best be describes as a celebration of corporate marketing and industry directed and funded “paint by numbers” research. Although a healthy majority of retina specialists in the trend survey use Avastin over Lucentis, a huge majority of presentations focus on expanding usage of Lucentis. Why is this? Is it because Genentech will support Lucentis studies and researchers but not Avastin studies? They will even offer free Lucentis and support to any Avastin study willing to switch. The sequence is to promote Lucentis research, support reporting that research at sponsored meetings, and then sit back while KOLs and, hopefully as many lemmings as possible, claim that we must use Lucentis because “that’s what the study was done on”. It is untenable that even as our “leadership” meets with Genentech to force the issue on this one question of Avastin availability, that they simultaneously organize the yearly marketing party for the companies that would like to control our every move.

New Imperatives-

So, what imperatives are we faced with? Many, it would seem. We clearly have a new standard of primary care in the treatment of exudative AMD. The AMD story serves as but one example of many, many moral difficulties that we are currently facing. We should re-evaluate and possibly suspend ongoing studies based on therapies that are clearly inferior to the new standard or that accept continuing vision loss as a “successful” endpoint. We should look at the design of new studies and define success endpoints based on what we know we can achieve in practice today rather than on the inferior standards set by PDT and Macugen. We should think hard about allowing research goals to be so strongly influenced by corporations. The alarming practice of corporate sponsors displacing local IRBs and serving as their own overseers should be banned entirely. Corporate participation of course is vital to medical progress, but this needs to be focused on public welfare and not only shareholder welfare. Pharmaceutical companies have a public responsibility unlike, say, a textile maker. They should be held accountable to the public interest and not be free to consider only their shareholders’ interests. Perhaps this needs to be regulated by better independent or governmental oversight. Finally and perhaps most importantly, we need to take more responsibility for the financial as well as clinical well being of our patients. Too many retina specialists have rationalized using expensive and sometimes inferior therapy by arguing that ‘the government is paying for it anyway’. When we have good data that a vastly less expensive off-label therapy is as good or better in terms of safety and efficacy, that is irresponsible. The system is our system, and if we don’t respect it, or worse, if we actively or passively assist corporations in manipulating the system for their financial benefit, then we have hurt our patients and the system in which we must all co-exist. The potential market cost of Lucentis used according to the RCT protocol is around 10.1 BILLION dollars per year for new AMD cases alone, and probably over three times that including other applications, when the budget for all of eye care is 4.5 billion and scheduled to decrease to 2.6 billion over the next 4 years! With the emerging discipline of evidence-based medicine, we have useful tools that will help us to demonstrate what therapies are most effective in our patients’ own estimation, and also which are most cost-effective. Time-tradeoff techniques and QALY based effectiveness analysis are emerging as powerful tools that will place in stark perspective the actual value of therapies that are marginal in terms of effect or cost compared with those that are vastly more effective and cost-effective. Current recommendations such as passage of the Sunshine law, pooled funding for CME and research, and a return to more humble, physician-centric, and sponsor independent society functions are vital first steps. The climate has changed forever, and it is time that we accept this and adapt.