Pages

Thursday, June 13, 2013

FDA's CFS/ME Drug Development Workshop: Day Two, Afternoon Panels

The focus of the first afternoon session was on designing clinical trials for drugs to treat CFS/ME, drawing on the experience of CFS/ME researchers and physicians.

You can read a summary of Day One, April 25th, HERE.

You can read a summary of Day Two: Morning Panels HERE.

You can read a transcript of the April 26 sessions HERE.

Panel 3: CFS and ME Clinical Trial Endpoints and Design (158:05) (VIDEO)

Moderators: Jordan Dimitrakoff, M.D., Ph.D., Assistant Professor, Tufts University, Boston, MA and Edward M. Cox, M.D., M.P.H., Director, Office of Antimicrobial Products, OND, CDER, FDA 

Speakers: Dr. Peter Rowe, M.D., Professor of Pediatrics, Johns Hopkins University School of Medicine, Director, Chronic Fatigue Clinic, Johns Hopkins Children’s Center
Christopher R. Snell, Ph.D., Professor, Health, Exercise and Sport Sciences, University of the Pacific
Elizabeth R. Unger, Ph.D., M.D., Chief, Chronic Viral Diseases Branch, Division of High-Consequence Pathogens and Pathology, Centers for Disease Control and Prevention 
Ashley F. Slagle, MS, PhD, Oak Ridge Institute for Science and Education (ORISE) Fellow, Study Endpoints and Labeling Development Staff, ONDIO, CDER, FDA (contractor)

First Speaker: Dr. Peter Rowe, “Clinical Trial Design in CFS.” [timestamp 7:10 – 48:09]
Peter Rowe

Dr. Rowe drew upon his experience over the past 20 years as a CFS researcher and clinician. He addressed several important issues pertaining to clinical methodology: what has and has not worked in clinical trial design, ideal patient populations, ideal endpoint, and potential study designs that might decrease the confounding effects of the heterogeneity of the illness, and which might help identify an effective therapy against the background noise of other comorbid conditions.

Main themes of Dr. Rowe's talk: 
  • Heterogeneity: Because CFS is so heterogeneous with variable onsets, and types of symptoms, as well as different combinations of comorbid conditions, heterogeneity needs to be taken into account when designing a trial for effective therapies. Design problems are the following: Careful selection of groups and subgroups, including those who recently have contracted CFS versus those with established symptoms, how to include patients with different levels of severity. And as with any illness with a lot of heterogeneity, small studies are almost certainly doomed to failure from the outset. We will need large studies to identify clinically significant but modest differences between patients.
  • Outcome measures that are currently in use are fairly serviceable. Imperfections in measurement tools are not the biggest impediment, but outcome measures need to remain simple.
  • Trial design: If trials are designed properly, they can yield clinically useful data.

Success and failure of previous clinical trials: Some historical lessons

The first of the modern trials was Strauss’ acyclovir study in 1988 (“Acyclovir treatment of the chronic fatigue syndrome. Lack of efficacy in a placebo-controlled trial.”). It had been hypothesized that CFS was due to a persistent Epstein-Barr viral (EBV) infection. In this study each patient received either acyclovir or placebo. Outcomes included measures of daily energy, how people felt, temperature, and mood. Of the 24 who completed the study, 21 rated themselves as improved, but improvement was evenly distributed among the placebo and acyclovir. (As an aside, Dr. Rowe noted that this was one of the few CFS studies in which there was a high placebo response. Usually CFS patients exhibit a low placebo response.) 

In spite of its failure to prove that acyclovir was an effective treatment, Dr. Rowe noted certain design features in the acyclovir study that should be used in current trials, namely crossover design (patients acted as their own controls by taking both the drug and the placebo), culling the study population to increase the likelihood of enrolling those who would most benefit from the intervention, insisting on a clinical evaluation to confirm diagnosis, and simple outcome measures. Although this trial was not a success, it revealed that while a small study could not determine what was effective, it could reveal the negative effects of medications. (Some measurements were actually worse during the acyclovir phase of the trial. i.e. mood and wellness). This is valuable information for future studies.

By 2006 there were 56 randomized trials and 14 non-randomized clinical trials for CFS. These included behavioral interventions (e.g. GET and CBT), immunological treatments (e.g. acyclovir, IVIG, imunovir), corticosteroids (e.g. Florinef, hydrocortisone), various pharmaceuticals (e.g. galantamine, NADH) and complementary therapies (e.g. massage, carnitine, liver extract). The outcome of nearly all these studies has been that pharmaceutical interventions are not helpful.

Why haven’t these trials shown that drugs can be effective therapies for CFS? Dr. Rowe pointed out that the main problem with these studies is methodology. Unless the patients share the same type of onset (sudden or gradual) and similar comorbidities (such migraines, IBS, allergies, orthostatic intolerance, to name a few), the studies run the risk of comparing apples to oranges. 

How can heterogeneity be reduced?
  • Careful subject selection. Clear case definition and clear eligibility criteria need to be established for subsets under study. If subsets are not identified, a predominant comorbidity, for example, can skew the results.
  • Identifying comorbid conditions. Flares in comorbid illness have the potential of obscuring treatment effects, especially in small samples. These fluctuations can overwhelm the effect of the treatment under study.
  • Large sample sizes. Given the heterogeneity of CFS and OI, large sample sizes are required, unless there is a powerful medication that applies across all subgroups.
Dr. Rowe stressed the advantage of large sample size by pointing out that while many CFS/ME studies were done using between 12 and 35 patients, the fibromyalgia study of pregabalin enrolled 529 patients. The outcome was simple: pain. Using a daily pain scale, they were able to show that 21% had more than 30% improvement in pain. This study shows that a large sample allows researchers to find results that are clinically significant for a small population.

Dr. Rowe also discussed the PACE trial in terms of its methodological success (with the quip that anyone who ventures into a discussion of the PACE trial probably needs an armed guard). Because the size of the trial was so large – 641 participants – the study showed a small but statistically significant difference (3 points). To put this in perspective, Dr. Rowe said that as physicians, a three-point difference in the Chalder Fatigue Scale (33 points) over a year would be enough to make them hang up their licenses and quit. But, he added, whatever our views are on CBT, it is important to focus on the methods lesson: the study was positive based on the sample size.

Dr. Rowe went on to discuss the Florinef trial conducted in 2001 (“Fludrocortisone acetate to treat neurally mediated hypotension in chronic fatigue syndrome: a randomized controlled trial.”). The study was based on Dr. Rowe’s clinical observations of several patients who had made startling improvements while on Florinef.

The study concluded that there were no significant differences between placebo and Florinef. However, when Dr. Rowe re-examined the results, he found that younger patients and those with a shorter duration of illness had responded to the drug, while patients who had been sick longer than three years had not. The failure of the Florinef study was due to lack of selectivity in patient selection. Additionally, Dr. Rowe pointed out that the success of the PACE trial may have been in part due to selecting for a patient group which had not had a long duration of the illness (2.7 years average). He compared this to the rituximab study which used patients who had been ill for a much longer period of time (5.8 years average).

What is the ideal endpoint?

In his discussion of the rituximab study, Dr. Rowe pointed out that results need to be measured within the context of specific outcomes: cognitive improvement, activity, and functional measures. Simply asking patients how they feel compared two weeks ago is not an accurate measure of improvement, because patients will increase their level of activity as they get better, generating the same level of fatigue. “We’ve got to watch that we don’t use fatigue as the only outcome measure.”

What are the best trial designs for CFS/ME?

Dr. Rowe’s recommendations for improving CFS/ME trial designs were as follows:
  • Larger study groups
  • Stratification for subsets
  • Run-in periods prior to the beginning of the study in which co-morbid conditions are treated and stabilized
  • Randomized trials in which ostensibly effective treatments are withdrawn (Ampligen)
  • Crossover designs, in which the patients take both the drug and placebo with a wash-out period (off the drug) to limit prolonged effects
  • Trials using a single patient

Christopher Snell
Second speaker: Christopher Snell, “Repeated CPET Results as Clinical Endpoints for ME/CFS Research.” [timestamp 48:36 – 85:42]

Dr. Snell reported on the value of cardiopulmonary exercise testing (CPET) to establish endpoints in CFS/ME studies. The beauty of exercise testing is that it measures function. Before CPET, nobody had an objective measure of fatigue. Simply asking people how tired they are is not real science. Dr. Snell believes the exercise test is a biomarker for CFS/ME because it provides both diagnostic and prognostic data.

The definition of fatigue is reduced efficiency as a result of doing work.

People are energy-producing machines with a built-in ability to stretch that energy when needed through the fight or flight response. Disease states reduce this capacity. But, even in disease states, in the absence of stress, reduction in functional capacity isn’t always seen. Exercise is an effective way to induce stress. Once stress is created, the true functional capacity of an individual can be objectively measured.

The goal of exercise testing is to assess the function of the cardio-respiratory system.

The way we produce energy is through the cardio-respiratory system, so we need to look at how efficiently that system operates. The cardio-respiratory system determines your functional capacity – or how much work you can do. This system functions on aerobic capacity – that is, how efficiently is the body able to use oxygen?

There are two main energy liberation systems: aerobic and anaerobic.
  • Aerobic metabolism – dependent on oxygen, is very efficient, can be used for extended periods of time, predominates at lower workloads, produces CO2
  • Anaerobic metabolism – no oxygen needed, predominates at higher workloads, 2ATP per glucose (as opposed to 36), produces lactic acid. Lactic acid is associated with pain, reduced muscle function, altered enzyme activity, cessation or reduction of activity.

Quantifying aerobic capacity 
  • VO2 Max is peak oxygen consumption. This is the maximum amount of oxygen a person’s system can deliver in order to produce energy.
  • Anaerobic threshold: the point at which a person switches from aerobic to anaerobic metabolism.

Both VO2Max and anaerobic threshold can be measured directly using gas exchange techniques and measuring blood lactate levels. Indirect measurements include heart rate and fatigue onset. Indirect measurements are not reliable. They don’t apply in disease states. Indirect measurements, such as field tests (e.g. the 6-minute walk) are also highly subject to bias. They can be influenced by motivation, the attitudes of the testers, and other factors.

As a case in point, Dr. Snell offered an interesting analysis of PACE trial. In the PACE study the 6-minute walk field test went from 341 at the beginning of the study to 414 yards a year later. Dr. Snell calculated the mph (miles per hour) as 1.9 at the beginning to 2.3 mph at the end. According to tables used to measure cardiac capacity, walking 6 minutes at 2 mph translates as severely disabled. Dr. Snell does not consider “severely disabled” to be an adequate endpoint for a trial, especially after 52 weeks of graded exercise. Because the improvement was so small, he attributes it to an increase in motivation.

How VO2max is calculated

Aerobic metabolism burns oxygen (O2) and releases carbon dioxide (CO2). Once CO2 exceeds O2, the person is approaching his or her maximum capacity. Aerobic metabolism is strongly dependent on the heath of the circulatory system. Anaerobic threshold is measured using gas exchange. If the ratio exceeds more than 1.1, the person’s CO2 is exceeding O2, which means that person is getting close to his or her limit. 

Because breathing is involuntary, this result cannot be faked. Gas exchange is therefore a much more accurate measurement than field tests.

Workwell test results for CFS/ME patients

Dr. Snell’s group was not able to distinguish CFS patients from controls until they looked at the results of the exercise. People with CFS were sick after the test. Strikingly, their ability to utilize oxygen decreased on the second day of testing. Controls were able to utilize oxygen better on the second day. Later, this study was replicated using 56 patients. While the results were not as dramatic, the second larger study showed that there was a big drop-off in efficiency. CFS patients produced less work for more effort and had a greater buildup in lactate.

Summing up

Post exercise morbidity distinguishes deconditioning from CFS.  CPET is a valid and accepted form of testing - there are just over 400 clinical trials around the world that currently include CPET (clinicaltrials.gov) - and should be used as an endpoint to evaluate the efficacy of any drug treatment in clinical trials.

Elizabeth Unger
Third speaker: Dr. Elizabeth Unger, “Measures of CFS in a Multi-site Clinical Study.” [timestamp 86:04 – 107:43]

[For additional coverage of Dr. Unger's talk, see the HealthRising article by Simon McGrath.]

Dr. Unger talked about a new CDC study that is currently gathering data to help define subgroups in CFS/ME. The study utilizes patient data from seven participating clinics: Drs. Natelson, Lapp, Bateman, Peterson, Klimas, Kogelnik, and Podell. Specifically, the CDC study addresses the main characteristics of heterogeneity: duration of illness, severity, comorbid conditions, medications and demographics. This study differs from previous CDC studies in two respects 1) it was designed to capitalize on the experience of the clinicians who treat CFS/ME, and 2) it collects standardized data to evaluate the heterogeneity of CFS patients across practices in order to revise the case definition and define subgroups.

The data gathered from the CDC study were generated via a physical exam, patient questionnaires – pain, sleep, depression and anxiety scales, CDC symptom inventory, and the DePaul symptom inventory (DSQ) – data extraction from medical records, patient illness history, tests, family history, and infection/immunization history.

First analysis of data from 393 patients in seven clinics (with variation between clinics):

Overall demographics:
  • Mean age 48.6 yrs
  • 71% female
  • 95.4% white 
  • Mean BMI 27.2 (overweight) [Calculate your body mass index HERE]
  • 58.1% married (16.1% previously married, 25.7% never married)
  • 78.3% college educated
  • 97.8% insured
  • 75.4 % not working (15.4% with unemployment benefits)(this was the only statistic that did not vary between clinics)
  • Illness onset
  • Mean age at diagnosis 38.2 years
  • Sudden onset 66.7% (with variation between clinics)
  • Mean duration of illness 15 years
Dr. Unger stressed that the patient population in this study is not necessarily representative of the CFS/ME population as a whole. For example, the vast majority of patients were white and insured, which does not correlate with demographic findings of Jason et al., in which CFS/ME was shown to have a high prevalence in non-white populations. Dr. Unger pointed out that this is an indication of lack of access to care that needs to be addressed.

Symptoms results

Fatigue scores varied not only between clinics, but between various instruments used to measure types of fatigue. Of note was the fact that using the general fatigue scale, 37% of the participants scored at the maximum range.

The SF-36 questionnaire was used to measure function. [Note: The SF-36 is a health survey with only 36 questions. It measures functional health and well-being as well as physical and mental health. It is a generic measure, used to compare the relative burden of diseases, and to compare the benefits produced by a wide range of different treatments.]

  • Interestingly, the highest scores on the SF-36 were mental health and emotional role (activity limitations due to emotional problems). [Note: This calls into question the frequent assertion that depression is a common comorbid condition of CFS/ME. It also reveals one of the flaws of using a case definition for CFS/ME that does not exclude clinical depression.]
  • The lowest scores were vitality and physical role (activity limitations due to physical health)
  • In this group of patients, daily vertical activities (upright or sitting with feet on the floor) averaged 7.2 hrs and the exercise average was 3.4 times a week, which means this was not a severely ill population.
  • Measures of pain: 80% had pain last week
Symptom scores were very illuminating [timestamp 101.50]
  • The symptoms that were most often reported as most severe were fatigue after exertion, and unrefreshing sleep 
  • The least: joint pain, sore throat, tender lymph nodes, fever, depression.  
  • Cognitive problems varied between those who had severe problems or none.
[Note: The distribution of symptoms is reflective of a patient population that is 1) older, and 2) has already passed the initial stage of the illness, in which flu-like symptoms predominate.]

Finally, Dr. Unger showed a chart that compared findings of the PROMIS scores for symptoms of CFS/ME against five other conditions: chronic pelvic pain, spinal cord injury, muscular dystrophy, post-polio syndrome, and MS. While sleep and pain that influences what you can do is about the same in chronic pelvic pain, in all the other measures: fatigue, sleep disturbance, pain, CFS/ME scored higher than the other conditions.

Conclusions: There is heterogeneity in the CFS population as a whole and measures based on symptoms alone are limited in their ability to determine subgroups. The CDC data will, however, allow the CDC to evaluate how well these instruments work.

Fourth speaker: Dr. Ashley Slagle, “Clinical Outcome Assessments to Evaluate Treatment Benefit in Clinical Trials for CFS and ME.” [timestamp 108:24 – 138:39]

Dr. Ashley’s main point was that in order to find effective therapies we must be able to assess those therapies in a reliable fashion. The FDA requires that drug developers provide documentation of “substantial evidence from adequate and well-controlled clinical trials” and that the methods of assessment of the subjects’ response be “well defined and reliable.” Well defined and reliable are the key criteria by which the FDA judges outcome assessments.

How the FDA measures treatment benefit

Benefits are determined by how patients feel, function or survive. Clinical studies use biomarkers to determine the effectiveness of a medication, however, biomarkers do not tell us how a patient feels, functions or survives.

Biomarkers may, however, provide indirect evidence of treatment benefit, by showing a biologic treatment response to a drug. Because there are currently no agreed-upon biomarkers to use as outcome assessments for CFS/ME, it is critically important to find assessments that can tell us how patients are feeling or functioning.

To determine how patients feel and function, clinical outcome assessments (COA) may be provided by patients, clinicians or caregivers. Unlike biomarkers, COAs are influenced by patient choices, because they depend on reporting. They require rigorous development and evaluation. Assessments reported by patients can also be used for epidemiology, prognosis, and other contexts. COAs do not need to be used in conjunction with a biomarker – FDA says that any assessment can be defined as well reliable.

Another important consideration when measuring treatment benefit is the context of use. Context of use includes disease definition (which must be explicit and specific) as well as characteristics (severity of disease). What is important is that the assessment match the population in which it is being used, taking into consideration age, severity, comorbidities, etc.

Challenges to developing outcome assessments: Subpopulations

One of the biggest challenges in CFS/ME is that the disease definition is not entirely clear. Therefore, it is important to identify a rational set of clinical trial entry criteria which permit the exclusion of other comorbidities as much as possible.

Another challenge is that subpopulations need to be defined. Some of these are:
  • Acute and gradual onset
  • Patients with OI and those without
  • Adults and children
  • Those with abnormal neurological findings and those without
  • Recent onset and those with long-time suffering
  • Severe forms and less severe forms
  • Patients with varying symptom experience
  • Others?
  • Defining subpopulations for clinical trials
There are already a number of existing instruments that can be used to measure the symptoms and domains of subpopulations of CFS and ME. But they have not been specifically developed for CFS and ME, so they may need to be modified to use for CFS and ME. (slide)

Some concerns are that some instruments have been developed for diagnosis, or epidemiology, so they are not useful for clinical trials. Generic measures are not specific enough, and include items that are not relevant to the CFS/ME population. There are conceptual framework concerns, for example how is “fatigue” defined? Sometimes the measures do not include activity level, which is very important in CFS and ME.

In spite of all the hurdles of complying with FDA regulations for determining the efficacy of a treatment, Dr. Slagle closed on an optimistic note. While none of the assessment instruments have been appropriate for CFS and ME overall, there are other options: assessment instruments can be used for specific subgroups, modified, and new instruments can be developed. To ease this process and ensure their conformity to FDA regulations, modifications can be made by independent groups (including those other than drug companies) and submitted to the FDA for use in future trials through the Drug Development Tool (DDT) qualification program.

Question and Answer Period: [timestamp 138:52]
Question: It’s clear that we need large clinical trials, because of the heterogeneity in the patient population, but those trials are usually paid for by pharma, which is not interested in funding large trials for CFS and ME. Patients do not have the resources to bridge this gap. Comments?
Answer: Dr. Rowe: This is a big problem. If studies don’t come from pharma, I don’t know who would fund them. Dr. Snell: This is a major economic problem. Yes, we need something to drive it. Dr. Slagle: If you develop good outcome assessments for pharma to use and lower the risk, and if they have well-established measures they can use, in consultation with the FDA, they are more likely to engage in clinical studies. Ms. Unger: Once we get a critical mass of data, and infrastructure, pharma will come. CDC is making a start on the data. Dr. Kogelnik: One of the promising funding avenues is a shared grant mechanism through a partnership with the NIH.
Question: Only 18% of patients with ME/CFS have those diagnoses exclusively, with no comorbidities. The high rate of comorbidity may be an additional challenge for clinical trials. Comments?
Answer: Dr. Rowe: Fibromyalgia has a high rate of comorbidities, yet it has a number of approved drugs. A large enough sample size will randomize comorbidities. Another option is that while any two people have a lot of differences, each person can act as his or her own control.
Question: Dr. Unger, infection/immunization is in your database. Have you made any connection between immunization and onset or relapse, and are the data maintained in such a way that such an analysis could be done?
Answer: Dr. Unger: We don’t have onset correlated with infection or immunization. We will have that information when the data are analyzed. We think this will be an important data point.
Question: Does the FDA assist with the development of trials and with developing outcome measures?
Answer: Dr. Slagle: There are two processes for outcome assessments. During the course of normal meetings with drug developers, the FDA will respond. The other option is the independent qualification process (DDT program) in which the FDA will provide specific consultation. The FDA welcomes the opportunity to talk with those who are designing trials and endpoints; the earlier the better.
Question: What does substantial improvement in VO2 max – and in steps walked – consist of (as per the Ampligen trial)?
Answer: Dr. Snell: We don’t have a lot of data related to treatments, because we don’t know what is causing the deficit. If the immune system is causing the problem, then treating the immune deficit will show improvement. There are no criteria for steps. Counting steps is not necessarily helpful unless you decide to relate it to something else.
Question: Is there insurance coverage for exercise testing?
Answer: Usually, insurance reimburses.
Question from Dr. Grobstein to Dr. Unger: The FDA says the patient population must be well defined for clinical trials, yet the CDC study does not ask participant clinics to use any particular definition. How do you reconcile these two views of how to do trials?
Answer: Dr. Unger: Ours is not a treatment trial. We are collecting data on what ME/CFS looks like in the clinic. That objective data can be used to develop a case definition. There are times when you want a broad case definition, and times when you want a narrow one.
Question from Dr. Grobstein to Dr. Slagle : You briefly pointed out problems with outcome measures – the instruments are not ideal. Does this mean we can’t do trials until we modify existing instruments or develop new ones.
Answer: Dr. Slagle: Not having great instruments shouldn’t hold up progress on trials. The problem is that although there may be a treatment benefit that’s there, it may not be picked up by the instruments being used. On the other hand, if there is a huge treatment effect, then existing instruments will show it. Until existing instruments can be modified, mapping the instrument to the population, even if it isn’t ideal, is what is important.
Question: What knowledge may be obtained by brain mapping, and what impact will that have on general research?
Answer: Dr. Unger: A lot of data will be coming from brain studies, eventually.
Question: What is the prevalence of CFS in different racial and ethnic groups? 
Answer: Dr. Unger: CFS is very common in racial and ethnic minorities.

Panel 4: Roundtable Discussion -- Summary and Path Forward (95:57) (VIDEO)

Moderators: Dennis Mangan, Ph.D. and Badrul Chowdhury, M.D., Director, Division of Pulmonary,
Badrul Chowdhury
Allergy, andRheumatologyProducts, ODE II, OND, CDER, FDA

Panelists:
  • Lily Chu, M.D., M.S.P.H. 
  • Jordan Dimitrakoff, MD, PhD 
  • Nancy Klimas, M.D. FACP, FIDSA 
  • Nancy Lee, M.D., Deputy Assistant Secretary for Health;Director, Office of Women’s Health, Department of Health and Human Services (HHS) 
  • Susan Maier, Ph.D., Deputy Director, Office of Research of Women’s Health, National Institutes of Health (NIH) 
  • Theresa Michele, M.D. 
  • Robert Miller, Patient 
  • Jody L. Roth, MS, RAC, Director Regulatory Affairs, Biomedicines Eli Lilly and Company
Question 1: What were the key messages on drug development you heard in this meeting?
Dr. Lily Chu: Data is necessary to do research. But not enough money is being spent on CFS research. Severely ill patients need to be studied.
Dr. Dimitrakoff : The key message is that it is actually possible to have a drug for CFS. The second message is that there is a great deal of support from the FDA supporting the development for medications for CFS/ME. This meeting validates how important it is to get young people involved in CFS research in the early stages of their careers.
Dr. Klimas: We are much further along than we think. Investigators are already linked together, and we have many years of experience using the instruments we’ve been talking about.
Dr. Nancy Lee: Partnerships with academia, patients, clinicians, pharma, and foundations are necessary to bring these efforts to fruition.
Dr. Susan Maier: Measurement is the key to making sure we have accurate representation of what happens in clinical trials and research.
Robert Miller: There is a crisis in the ME/CFS community and that crisis is lack of treatment. Measurements already exist that can be used to evaluate CFS symptoms. The government needs to take action now to clarify a very clear special pathway to approval for drugs for ME/CFS. This will make it clear to pharma that FDA will support drug development. Fund it and they will come. Treatments must not be withheld because of the heterogeneous population. FDA should move within the next two months to create a set of criteria and outcome measurements. CDC and NIH should study the responders to Ampligen to find out for whom it works and why. This is a crisis, this is an emergency, this is a time for everyone to work together and try to move forward.
Dr. Theresa Michele: Patients were articulate and clear.
Jody L. Roth: It is essential for pharmaceutical companies to understand endpoints. We have the correct framework to continue forward.
Question 2: What do you think are the most important factors in facilitating drug development in CFS and ME? [timestamp 29:21]
Jody L. Roth: What are the criteria we need to have in place for clinical trials? What is the framework, how are we going to implement the trials to get the endpoints in place? How can we leverage other fatigue instruments we have used in order to find out if our drug will work in this syndrome?
Dr. Theresa Michele: We have a need for data. We already have a lot of data, but we need more longitudinal data. There are a number of different databases already out there. We need to establish networks to bring those databases together.
Robert Miller: Willingness from all the federal health agencies to work together and figure out the pathway to do clinical trials. A plan needs to be in place in order to move forward.
Dr. Susan Maier: Be willing to accept researchers working on other diseases to broaden how we do research.
Dr. Nancy Lee: Developing data infrastructure is very important. Drug repurposing has promise in the shorter term in order to find useful therapies that have already been approved
Dr. Klimas: The most important facilitating factors have to do with organizational infrastructure. We have the platforms, and things are cheaper than they used to be. It is very reasonable to think about putting together a series of clinical trials. Subgroups of patients can be a strength, because it helps us target specific drugs and therapies. There is no reason, at this point, in delaying any further. We have good reliable, validated, well-published variables already. We’re ready to go.
Dr. Dimitrakoff: The key word is collaboration. It is important to work together with people in similar and related fields.
Dr. Lily Chu: A big piece of research is educating researchers and physicians about CFS/ME. 85% of physicians think that the illness is psychiatric. There are biomarkers for CFS/ME. [applause] What’s more CFS is not the only disease with a lot of heterogeneity or comorbidities. It is not a unique disease.   
Question 3: Based on the discussion from Panel 3, what clinical trial design elements are most important to ensure success of drug development programs for CFS and ME? [timestamp 43:42]
Dr. Nancy Klimas: Important design elements are: sample size, symptoms that have to ability to improve and which you can measure, and the length of the trial itself. Little 10-week and 4-week trials may not be enough. You need at least four months to get even a hint of efficacy. Our group is using an exercise challenge to induce relapse, intervening with the targeted drug, and then challenging the system again. This not only demonstrates the efficacy of the drug, but it teaches us more about the dynamic of the illness.
Dr. Michele: It is critical that we not forget measures of safety. Efficacy is only half the equation. So we need to understand clearly what the risks are, even for drugs that are being repurposed.
Dr. Lily Chu: There are two papers which might be interesting to pharmaceutical companies. One is by Haywood, “Quality and acceptability of patient-reported outcome measures used in chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME): a systematic review.” [Read the abstract HERE.] The other paper is by Cockshell, “Test effort in persons with Chronic Fatigue Syndrome when assessed using the Validity Indicator Profile.” [Read the abstract HERE.] 
Question 4: What do you think are the most important barriers to conducting research for CFS and ME, and what can be done to overcome them? [timestamp 47:36]
Dr. Michele: From the FDA ‘s perspective, we don’t perceive any barriers. We have outcome measures, we have definitions, we have symptoms that can be used. Regulatory-wise we are ready to go.
Dr. Chowdhury: The time has already arrived to enroll patients in trials.
Dr. Dimitrakoff: There are two barriers that have been consistently pointed out: lack of new investigators coming into the field, and limited availability of funds. All the clinical centers should come together and do a large clinical trial. The NIH has something called the U34 grant, which allows teams of investigators to come together to design a clinical trial over a period of one or two years. Following that period, there is an additional funding period of five years when you actually do the trial.
Dr. Maier: Money is the primary obstacle. What drives funding from the NIH is well-conceived ideas, good science and well-written proposals. We can’t fund research, even requests for applications, without a proposal. [Submit them and they will come.]
Question 5: How can we best leverage your individual experiences in order to facilitate drug development in CFS and ME? Please respond for your group (Health and Human Services, FDA, pharma, academia, patient/advocacy).
Dr. Michele: FDA can express its willingness to work with companies and investigators on drugs for CFS. If FDA considers CFS to be a serious disease, this sends a message to pharma that this may be a good opportunity to make some money. Money makes the world go round. When people come in with clinical trials, we can help them with trial design, and with lessons learned from other diseases. Our division has its doors wide open for applications for CFS.
Dr. Maier: Call me. [This was a direct personal invitation to researchers to seek advice and recommendations on proposals from Dr. Maier.]
Jody L. Roth: Pharma needs to get together with these consortia to establish endpoints and measurements that we can use.
Dr. Lee: We need data infrastructure, and the CDC is doing that.
Bob Miller asks Dr. Klimas to talk about her applications.
Dr. Klimas: We write eight applications for every grant that we get. From the point of view of academia, we need philanthropic support. You want us to be quicker, so that you can have your drugs, but I am struggling to get Phase I work done. It’s not that we don’t have good ideas, or good science. What we don’t have is your time to waste. Dr. Klimas stated that she has nine grants on the board right now.
Dr. Chu: IACFS can help disseminate information. Patient groups are enthusiastic to work with anyone. NAAME is a patient organization that will help facilitate this. http://naame.org/
Question 6: What are possible next steps following this meeting? (Please respond as per your group.)
Dr. Klimas: We need to have a small meeting with the experts in the FDA to talk about outcome variables, and really nail them. Dr. Klimas volunteered to come back to Washington DC to do that. (Dr. Dimitrakoff did as well.)
Jody Roth: we need regulatory effort for trials to move this forward.
Dr. Michele: Guidance for industry is the next step. This will not happen quickly, but we are very well aware that it needs to happen.
Dr. Dimitrakoff suggested that a peer-reviewed paper should come out about the FDA meetings, so that the larger medical community has some idea that the field is moving forward. [applause]
Dr. Chu: Publishing in a peer-reviewed journal is also a way to engage new researchers.
Bob Miller
Question and Answer Session [timestamp 74:10]

Questions: Are there clinical FDA trials approved for CFS? Why was Ampligen not approved? Does FDA have a working group to write industry guidelines?

Answers: First question: There are approved clinical trials, but not too many. Second question: FDA sends a response letter to the drug company and that is confidential. [Third question not answered.]

Question: Bob Miller: What are the approved clinical trials?

Answer: Dr. Michele: Just google CFSAC.

Question: How can repurposing of drugs be financially worthwhile for a pharmaceutical company?

Answer: Jody Roth: We are look for line extensions, or new uses, to add to existing drugs.

Question: Is anybody lobbying Congress to set aside money for CFS research?

Answer: Lily Chu and Bob Miller said they were both lobbying, but they did not have sufficient energy to do so.

Question: How are serious adverse events taken into account, especially in light of the fact that any hospitalization is considered a serious adverse event, whether it is drug-related or not.

Answer: Dr. Michele: This is why we need large groups for clinical trials. Serious adverse events can then be measured in both treated and non-treated groups.

Question: We have biomarkers, so why does the FDA continue to insist that we have none? What needs to be done to validate NK cell function, immune markers, and viral titers as biomarkers?

Answer: Dr. Chowdhury: Biomarkers are not needed for clinical trials. There is no up-front need to develop a biomarker. Dr. Klimas: The primary outcome needs to be function. What we need is something to predict function, and, yes, I do think we have the data to support that we have biomarkers that predict function. Both Dr. Chu and Jody Roth pointed out that biomarkers are indeed necessary for establishing subgroups and outcomes.