Megan Polanin, PhD, National Center for Health Research: June 15, 2017

Thank you to the FDA and the Safe Use Team for hosting this meeting and for the opportunity to share our perspective. I’m Dr. Megan Polanin, and I am a Senior Fellow at the National Center for Health Research.

Our focus is patient-centered comparative effectiveness research. With support from PCORI (the Patient-Centered Outcomes Research Institute), we’ve trained about 120 patient advocates to understand the value as well as the limitations of clinical trials.

Some of the patients we’ve trained are desperate for new cures, and others have been harmed by medical products and believe that the FDA needs to require better evidence of safety before approving new drugs or devices.

But the one thing they all agree on is that patients aren’t really getting informed consent about the drugs they are taking. Most doctors aren’t talking to most patients about potential side effects, or even about how good the evidence is that the treatment will work. Most patients blindly fill prescriptions because they assume that the drugs are proven safe and effective for patients just like them.

A big problem is that doctors rarely tell patients if their prescriptions are off-label. Most patients who are prescribed a drug off-label don’t realize that the drug they are prescribed has not been approved for the particular indication they need. Research shows that off-label uses are often ineffective and have more adverse side effects than on label use. Patients deserve to know if a doctor recommends a use that hasn’t been proven to have benefits that outweigh the risks.

Reducing the chances of adverse drug events requires a discussion between the doctor and patient about any evidence that the drug really will do more harm than good.  Unfortunately, doctors often have limited information – most of which they got from drug reps, not from the FDA. Drug labels have gotten so long that most doctors and patients don’t read them. I ask you: Why include the chemical composition of the drug on those labels? Why not focus on the risks and benefits in simple language instead?

And, as we all know, many doctors don’t have a lot of time to discuss these kinds of issues with their patients, and many doctors are not skilled in having these kinds of conversations.

In addition, FDA rarely provides information to patients or doctors about adverse events for specific major demographic groups, such as women, men, people of color, or for older patients. Having that information clearly on the label would help reduce adverse drug effects.

Doctors and patients should be clearly told when a drug is approved on the basis of biomarkers, and whether there are other drugs for the same indication that are proven to improve survival or clinically meaningful health measures, such as fewer heart attacks or better quality of life.

Bottom line: An important way for the FDA to help prevent adverse drug events is to make sure that patients know in advance what kind of choices they have for prescription drugs, and which ones are likely to be safer for them as women, people of color, patients over 65, and people with a particular illness. Patients deserve to know this information before filling their prescription.

Stephanie Fox-Rawlings, PhD, National Center for Health Research: May 24, 2017

Thank you for the opportunity to speak today. My name is Dr. Stephanie Fox-Rawlings. I am a Senior Fellow at the National Center for Health Research. Our research center analyzes scientific and medical data to provide objective health information to patients, providers and policy makers. We do not accept funding from drug or device companies so I have no conflicts of interest.

The pivotal study that is the basis of today’s review only demonstrated a small improvement in the primary efficacy endpoint. After 2 years, about 2.3% more patients were without invasive disease if they took the drug compared to placebo. This difference was statistically significant, likely because the large number of patients in the study. However, such a small difference could be specific to this particular sample of patients and trial and might not be generalizable for all women with early-stage breast cancer.  It is impossible to say, since after 2 years over 90% of patients were free of invasive disease whether they received the drug or placebo.

Patients followed for 5 years had a similar result – about 2.5% were more likely to be cancer-free, while almost 90% of patients taking placebo were also cancer-free.   There are no data yet on overall survival, so the results aren’t compelling.

This small difference should be considered in the context of adverse events that are typical of cancer drugs: diarrhea, nausea, vomiting, and fatigue were common. However, some were categorized as serious events. Adverse events were so unpleasant that they caused 28% of patients taking the drug to drop out of the study, compared to just 5% of patients taking placebo.

The sponsor also presented data from an ongoing, open-label, single armed study aimed to reduce adverse events due to diarrhea with prophylactic treatment. However, there were still a high occurrence of diarrhea, and the treatment for diarrhea caused a different set of adverse events.  

Patients should not be exposed to adverse events if the drug isn’t proven to provide a real improvement.   The 2.3% difference between 91.9% and 94.2% is not impressive.  And with only one pivotal study, there is no way to know if that result would be replicated in a second study.

A recent study published in JAMA Internal Medicine found that when FDA approved cancer drugs based on a surrogate endpoint, such as cancer-free survival, later studies have not found a benefit in overall survival.  And yet, these drugs cost an average of $100,000 – often more – and can harm quality of life.

We see that neratinib’s benefit compared to placebo is similar to that of a previously approved drug.  That does not mean that it should be approved. Patients do not benefit from more new drugs on the market unless the new drugs are more likely to have benefits that outweigh the risks.

The FDA should be sure that new treatments provide a real benefit to patients before they are approved.   We recommend that the FDA not approve neratinib for breast cancer unless a clear benefit can be replicated, or a benefit for overall survival is demonstrated.

The Cancer Prevention and Treatment Fund is the major program of the National Center for Health Research, and can be reached through Stephanie Fox-Rawlings at sfr@center4research.org.

Jack Mitchell, National Center for Health Research: May 9, 2017

Good afternoon.  I appreciate the opportunity to address this distinguished panel and our FDA participants.  I’m Jack Mitchell, director of government relations for the National Center for Health Research (NCHR).   We conduct research, use research data to inform public policy, and advocate for safe and effective medical products.  NCHR accepts no pharmaceutical or medical device industry money, and therefore I have no conflicts to report.

We strongly support FDA efforts to strengthen the role of patients and we urge the agency to define patients as those who use medical products, whether or not they are seriously ill.  The patient-focused meetings with FDA and other patient initiatives mentioned by Dr. Mullins in her presentation are welcome and necessary.  I’d like to speak to broadening those patient perspectives.

We know it is a challenge for FDA to attract patients who are truly independent, since so many patient organizations are funded by industry, and many patients are trained and recruited to participate in FDA meetings by industry.  Of course, all patients deserve to be heard, but  industry-supported perspectives should be augmented by independent patient voices.

For example, a recent study by Harvard researchers found that almost all patients who spoke at FDA public meetings had ties to companies that could benefit from their remarks.  Another study, of an FDA advisory committee meeting on a drug for Duchenne’s muscular dystrophy, reported that of the fifty-one public speakers, all but one had financial ties to the company that makes the drug.  That one public speaker was from our research center.

There are patient organizations that are not funded by industry and can offer a more independent voice.  We suggest that FDA needs to do more to reach out to them and include them.  For example, the USA Patient Network is a new national organization consisting of patients who have received training to help them understand clinical trial research design and analysis, so they can serve as confident, well-informed patient representatives on FDA and NIH advisory committees.

In addition, the Patient, Consumer and Public Health Coalition is an informal coalition of about two dozen non-profit patient, consumer, physician, and public health organizations.  They work together to prepare public comments for the FDA and other health agencies, and to educate Congress about important health policy issues.

Patients from these organizations have made presentations before FDA advisory panels and public forums.  They usually pay their own way to FDA meetings.  Not surprisingly, they are dramatically outnumbered by patients whose travel, meals and other expenses are reimbursed by industry.

We respectfully ask the Science Board to ensure that FDA enhances efforts aimed at including these independent patient voices, not only for new drug development, but also on the wide range of public health initiatives in which the agency is engaged.

Your focus today is on innovation initiatives mandated by the 21st Century Cures Act.  We are concerned that the new law does not guarantee sufficient resources to implement all its FDA-related provisions.  For instance, it encourages the FDA to rely more often on preliminary data such as biomarkers, and allow third party review to replace FDA’s premarket scrutiny.

The law has already resulted in the FDA deregulating more than one thousand moderate risk devices that will no longer be required to submit 510(k) applications.  To better ensure safety, FDA needs to expend more resources to improve post-market surveillance, especially of medical devices.

Unfortunately, neither 21st Century Cures nor the user fees that FDA has negotiated provides sufficient resources for effective post-market surveillance, particularly for medical devices.  Patients from the USA Network and other independent patient organizations have provided documented evidence to FDA of serious, irreversible harm caused by fast-tracked device approvals and inadequate post-market surveillance.  They tell us that FDA often is not sufficiently responsive to their requests to strengthen patient safeguards.

We respectfully urge the Board to carefully address these patient recruitment and safety issues as you advise FDA about implementing the 21st Century Cures Act.  Engagement perspectives should include patients who have been harmed by medical products that were not as safe as the research indicated, or had risks about which the patients were not adequately warned.  While FDA is appropriately and routinely hearing from patients desperate for new treatments, those are not the only patients who have important perspectives from which the agency can learn.  Thank you.

December 3, 2014
Dr. Anna E. Mazzucco

Thank you for the opportunity to participate in this meeting.  My name is Dr. Anna Mazzucco, and I am speaking on behalf of the Cancer Prevention and Treatment Fund.  I received my Ph.D. in cell biology from Harvard Medical School, and I conducted post-doctoral research at the National Cancer Institute.  Those are the perspectives I speak from today.

Our organization conducts research and shares information with health professionals, patients, and consumers.

We applaud the efforts of the EPA to protect the public from harmful chemicals in their food, water, air and products they use every day.  We strongly support the agency’s efforts to test chemicals more efficiently, and with the best possible scientific methods.  Research has implicated endocrine-disrupting chemicals in cancer, infertility, and other health problems.  We must be able to quickly identify such agents in order to prevent them from entering our food and environment, and from affecting our health for generations to come.

Use of high-throughput technologies is critical to expediting chemical screening, and we support the agency in these efforts.  However, we have several concerns regarding the high-throughput screening approach as described in the EPA White Paper.  Specifically,

• The EPA needs to spell out more clearly how prioritizing will be decided using this screening approach. The program should explicitly prioritize compounds based on risk contexts, such as those with widespread and persistent exposure profiles, and those with the most serious potential health effects.
• We know that chemical exposures during sensitive periods such as prenatal and adolescent development can greatly impact health in adulthood. These critical windows should be prioritized in bio-monitoring and population-based studies.  This approach would protect the most vulnerable among us, and also ensure the greatest public health benefits for everyone.
• Other stakeholders have expressed concern over lack of clarity regarding the relationship between the area-under-the-curve model and potency as indicated by traditional assays. We are also concerned that the model has not been tested on a sufficiently diverse set of chemical structures.  The agency needs to address these issues and provide additional evidence to support use of the current methodology, as opposed to other possible models.
• These screening assays must be relevant to real world exposures in order to be meaningful. Broad bio-monitoring and exposure studies are vital to informing extrapolation modeling in order to reap the benefits offered by high-throughput screening.
• As these high-throughput assays are further developed, the most orthogonal assays with the most downstream read-outs possible should be used, in order to capture the broadest possible range of mechanisms-of-actions.
• We agree with the EPA that metabolites could be missed using in vitro studies alone, and therefore some in vivo studies will be needed.
• We strongly urge the EPA to make the computational model fully accessible for independent assessment. That would increase transparency and stakeholder participation in this project.

Lastly, we share the concerns expressed in the Inspector General’s 2011 report that this process has not progressed as quickly as it should have.  We strongly urge that adequate resources be dedicated to these efforts, in order that the agency may meet its goals in a timely fashion.

Thank you for the opportunity to address the panel today.

October 16, 2014

Thank you for the opportunity to speak today.  I’m Dr. Diana Zuckerman, president of the National Center for Health Research.  I’m trained in psychiatric epidemiology at Yale Medical School, I’m a former faculty member at Vassar and Yale and a researcher at Harvard , and I’ve taught Research Methods courses, and those are the perspectives I bring with me today.  Our Center has no financial ties to Chantix or its competitors or to Chantix lawsuits.

We all know that smoking kills thousands of Americans and it is very difficult to quit.  That’s why we believe that Chantix should be available as an option for those who can use it safely.

At the same time, patients and their physicians need a clear black box warning for Chantix so they know to stop taking it when necessary

The challenge today is: which data should the FDA believe?  Mark Twain once said there are 3 kinds of lies: Lies, damn lies, and statistics.  I am a scientist and I believe in statistics, but I also know they can be easily manipulated to support a particular point of view.

Your task today is to make sense of conflicting data and decide which to believe.  They include:

• Meta-analysis
• Observation Studies based on hospital records
• Adverse Reaction Reports from physicians
• Reports from Patients

Meta Analysis is a valuable tool but its accuracy depends on the quality of each study and whether they fit together.  Meta analysis results can be useful or inaccurate depending on which studies you include and exclude.  No justification was given of why most studies on Chantix were excluded and only 5 were included in the meta analysis, including one study of schizophrenics, one study of depressed patients, and 3 studies of mentally healthy patients.  It is important to study schizophrenics and depressed patients, but those data should not be mixed together with 3 studies that exclude such patients.  No justification was given for that decision, but you heard from the FDA that most psychiatric events were in those two groups of mentally ill patients, clearly biasing the results.

To consider the studies showing no association between Chantix and psychiatric side effects, it is important to understand what happens to people with acute psychiatric events related to medication.  As the FDA speakers pointed out, most do not end up in hospitals or the ER.  Many of these psychiatric side effects are not reported in medical records.  Because psychiatric commitment laws depend on acts of violence, not threats of violence, many people with dramatic psychiatric symptoms end up in jail, not in hospitals.  In fact, some studies show that there are more mentally ill individuals in the criminal justice system than in psychiatric facilities — certainly those who suddenly behave violently toward others are likely to be put in jail, not in a hospital or ER.

There’s another, more positive reason why these psychiatric side effects might not be measured in a large study.  Fortunately, many stop quickly because patients or their doctors realize they should stop taking the drug, thanks to the black box.

For all these reasons, most of the studies that Pfizer is relying on are fatally flawed.

How can we make sense of the studies showing no impact in light of the thousands of reports of psychiatric side effects?

• The studies cited did not evaluate all psychiatric side effects, they focused on depression and suicidal thoughts and behaviors
• Those studies did not interview patients – a shortcoming of many large databases
• They relied on hospital records, which research shows missed 82% of adverse psychiatric events
• Some also relied on ER or medical records – which is better than hospital records, but will still miss a lot of data 

What about studies showing a significant increase in psychiatric events?  We all know that adverse reaction reports are the tip of the iceberg – it’s a voluntary system of reporting.  Compared to medical records, they can have a richness of information.  And while they are far from perfect, the sheer volume of thousands of reports – many more than for other drugs – is very compelling.

I’ve spoken with some patients who took Chantix, and their reactions are distinctly different from many other drug side effects, and don’t fit neatly into the categories that most of the Pfizer studies evaluated.  For example, I spoke to a man who was so besieged with uncontrollable thoughts that he locked his door at work and wouldn’t let anyone in.  His thoughts were so terrible that he just couldn’t deal with anyone.  That psychiatric reaction would be unlikely to fit into any of the studies.  Or a man who was so frightened that he hid in the corner of his bedroom under a blanket, trying to escape the uncontrollable thoughts by being as small as possible – trying to feel safe.  What study would accurate evaluate that?

If this Advisory Committee ignores the compelling psychiatric adverse reactions that have been reported, it would discredit thousands of doctors who made thousands of reports.  It would also discredit thousands of patients who reported them.  And it would send the message to the FDA to stop their Adverse Event Reporting systems, because what is the point of having such systems in place if you ignore thousand of such reports?

We need better studies, and I hope the post-market study underway will be better.  Based on previous research, we know that such studies must include very large numbers of patients, and must follow patients for a long enough time – not all reactions are within 30 days.  And, the studies must include patients’ reports of their side effects

In conclusion:

• The studies Pfizer is citing are fatally flawed because they omit most psychiatric adverse reactions
• Deleting the black box would send the message that thousands of doctors’ reports don’t count, including suicides and homicides

We strongly urge you to urge the FDA to keep the black box warning to protect patients and that you strengthen rather than weaken that boxed warning.  And, since the meta analyses are fatally flawed, as I and others have pointed out, the FDA should delete the misleading meta-analyses info from the Chantix label.

October 9, 2014

Division of Docket Management (HFA-305)
Food and Drug Administration
5630 Fishers Lane, Rm. 1061
Rockville, MD 20852

Comments of the Cancer Prevention and Treatment Fund on
“Methodological Considerations in Evaluation of Cancer as an Adverse Outcome Associated With Use of Non-Oncological Drugs and Biological Products in the Postapproval Setting”
Docket No. FDA-2014-N-0731

The Cancer Prevention and Treatment Fund appreciates the opportunity to comment on Methodological Considerations in Evaluation of Cancer as an Adverse Outcome Associated With Use of Non-Oncological Drugs and Biological Products in the Postapproval Setting.

Challenges

Due to its biological complexity and often long latency, cancer represents a challenge for monitoring in the pre-approval or postapproval setting.  Pre-market studies are usually too short-term or too small to adequately evaluate cancer as an adverse event.  Unfortunately, many postapproval studies are also too short-term, too small, or have too many patients lost to follow-up to accurately evaluate cancer as an adverse event.

Existing voluntary databases for adverse event reporting in the post-market setting are known to greatly under-report and often do not contain sufficient information.  Postapproval studies with pre-specified endpoints often have poor patient retention or are not completed.  While cancer registries can provide another source of information they often do not contain detailed drug history and are not nationwide.  Therefore, comprehensive systems for monitoring cancer as an adverse event for drugs and biologics do not exist at this time.

Long-term surveillance is critical

Cancer development is a multi-step biological process that can occur over several years, making long-term surveillance critical to identifying a cancer safety signal.  Short pre-market studies that are not specifically designed to assess cancer outcomes are unlikely to identify this potential hazard.  In addition, patient subgroups that may be particularly susceptible to cancer risks, such as children, adolescents and elderly adults taking multiple medications, are typically under-represented in pre-market studies, further decreasing the likelihood of detecting a cancer risk.

Post-market surveillance should consider all cancer types           

While concerns over specific cancer types may arise from pre-market studies, post-market surveillance should include consideration of all cancer types.  As any cancer signal can be difficult to detect pre-market, post-market surveillance needs to account for all possibilities.  As mechanisms of carcinogenesis are often shared across different tissues of origin, the possibility of cancer at multiple sites cannot be ruled out.

In 2006, the Government Accountability Office issued a report stating that the FDA needs to address weaknesses in its post-market surveillance system, including the need for larger and more diverse datasets using uniform systems such as electronic health record information.  That same year, the Institute of Medicine recommended that the FDA should take action to partner with other federal agencies and use all available resources to bolster its post-market surveillance efforts.  We strongly agree.  We recommend that the Sentinel System be used to mine specific data on the possible increased cancer risk due to the effects of non-oncological prescription medications and biological products, but Sentinel should not replace postapproval studies designed to evaluate those risks.

Concerns about cancer risks in pre-market development

If concern over a potential cancer risk arises during pre-market development, well-designed preclinical and clinical studies to directly examine this possibility need to definitively address this issue.  Specifically, sponsors need to have a clear understanding of the biological mechanism of action of their product, and perform comprehensive carcinogenicity testing.  This should include sensitized animal model systems as appropriate, and should also address non-genotoxic mechanisms of carcinogenicity, such as hormonal effects, which may display non-traditional pharmacological behavior, i.e. non-monotonic dose responses.  Any cancer concerns identified in pre-market studies should also be adequately addressed in labeling information, special warnings, and post-market studies as needed, if the product is approved.

Conclusions

To better evaluate the challenges in designing postapproval studies to determine whether a non-oncological drug causes or influences cancer, the FDA should develop long-term studies that include susceptible subgroups such as children, adolescents and the elderly.  The post-market surveillance should consider all types of cancer, and include large and diverse data sets utilizing electronic health records and the Sentinel System.  However, Sentinel should not be used as a replacement for postapproval studies.  Also, if pre-market studies flag a cancer concern, it should be addressed in the label of the drug with a special warning.

Cancer Prevention and Treatment Fund

The Cancer Prevention and Treatment Fund can be reached through Paul Brown at (202) 223-4000 or at pb@center4research.org