 Well, I've really enjoyed those parts of the media I've been able to be at and especially that last talk and following up on the radio analogy, I feel like since I left academic medicine that some of what I've been doing is working with biologists who are trying to fix radios and trying to keep them from having it blow up when they plug it in. And maybe the radio in this case is the patient. And we're all working to the same end, which is to really bring these amazing insights from biology into advances that can benefit patients. So I'm going to talk at a conceptual level, but most of the concepts, because this is a very thoughtful field in general, you've already heard about, so I apologize for that. But I'll sort of do it at that level going all the way from how do we go from the biology to the products and what are some of the observations that I've made in working in this continuum that might be helpful in working together. So I'm going to start with where I see as a non-microbiome person, although I've sat in a few meetings about microbiome and had many discussions about it, but how does it look from somebody who's not directly involved? From a clinical and FDA perspective, just some reminders that it's important to remember that sometimes what seems like it works really doesn't, and sometimes what makes sense isn't that sensible, and sometimes what looks safe isn't. Talk a little bit about, you've heard a couple of discussions that have touched on causation because I think our thinking about our observational science should drive what we move forward, and we need to be very careful about that. And there are some, it was really fascinating to me to think about this talk because there are some things about, obviously the microbiome isn't a drug or a vaccine, and when we think about manipulating it, the normal concepts applied to how we think about clinical development and clinical intervention need to be readjusted, or at least considered, and that leads to what are unique aspects of the products, points to consider as we do clinical development, clinical trials, and then in general, how can we work together to move this field forward? And I'm really impressed actually by the maturity of the presentations I've heard from basic scientists in terms of understanding sort of the connections that need to be made to have this work, i.e., a sophisticated understanding about causation, and a sophisticated understanding about interventions and how complicated they are. So in terms of where do things seem to be, while I really, as an infectious disease physician and former basic scientist, I'm thrilled that this is, I believe that the concepts of microbiome are really, are a breakthrough in terms of understanding what a human being is, really. We also see, as in so many areas of science, a huge data explosion triggered both by a conceptual shift but also by technology and bioinformatics. And we're seeing, I'm seeing this in so many areas and it's almost tried to say it, but we have so much information and yet we really don't have much understanding of most of that information, and one of the problems that I've been seeing is that there are less and less people who can actually understand the full implications of that information. So anyhow, the excitement and enthusiasm are legitimate. We've heard a lot about the gaps in understanding at this meeting. These include things like what really is normal and what isn't, what are the dynamics of change, what is the, how does the inventory of the microbiota relate to functional status? And I really like the concept, several people have advanced that it's really what are the inputs and outputs of this community that may be most important for the human host. However, we can't forget that some of these organisms may also function as traditional bacteria. They may trigger receptors. They may do other things beyond their metabolic products, so we don't want to forget that. We're seeing lots of disease associations and we don't know if they're causative. I'll get into that a little. Animal data, some of which you seem can be quite exciting and suggest that manipulating the microbiome can impact disease or risk factors. I won't go into the successful intervention in C. diff, but clearly there's some promising clinical observations and that's been one of the most, I would say, remarkable ones as a physician who has taken care of people with refractory C. diff disease. This is a real problem and this is a real breakthrough, so that's fantastic. There are other areas that sort of are less focused on, but utility, for example, in predicting disease, could we predict disease based on the microbiome component when we just heard that, in fact, the genome may drive the microbiome, which then may drive the disease. And can we utilize the microbiome to modulate response to drugs and vaccines or to better understand them? So I did want to mention a couple of the cautionary tales, more as a framework than as any negativity, but there are a lot of examples of things where the best scientists in the world are totally convinced that something works and physicians are convinced that something works and it doesn't. A big one, and I used to do transplantation, was autologous bone marrow transplantation for breast cancer. This got to the point that there were demonstrations that clinicians and patients were demanding payment for this procedure, which is basically getting your own bone marrow harvested, getting your entire bone marrow ablated, and then re-infusing your own bone marrow. It's a pretty toxic procedure with some finite mortality rate, and there were story after story of people with long-term survival and metastatic breast cancer that seemed to substantiate that this was beneficial. And finally, and I think this was a breakthrough in how we think about effectiveness in medicine, payers said, okay, we're going to pay for this, but we're also going to, I think CMS was involved, we'll do this in the context of also funding a study for those people who would be willing to enroll in a study. And a big controlled study was done, and lo and behold, there was no benefit of this procedure. So you had thousands and thousands of women get these autologous transplants, spend lots of time in hospital and bone marrow transplant units, and actually end up having more mortality and morbidity. So that's a cautionary tale. Also, this whole thing about markers that really seem to make sense to us don't always predict disease. So there's premature ventricular contractions, which are a sign of an irritable ventricle, which is a sign of predilection to sudden death, and is fairly predictive of it. But there are drugs that have been studied where that was chosen as the endpoint suppression of that ventricular ectopy, which is believed to then trigger sudden death, where that ectopy was suppressed, but actually sudden death increased. So again, the biology is sometimes more complex than we want to think. Now, when we get to biologics, which are either made from or of living things, obviously, as everybody in this room is where product safety becomes a particular concern. And things done with the best intention have had some problems. Fortunately, they are very few and far between. We have a, you know, generally a very responsible industry, and we have a strong regulatory agency. But actually, the entire FDA was pretty much founded because of a horse named Jim, as when it was realized early in the last century that plasma could be protective against infectious diseases. Horses were immunized against infectious diseases. And a diphtheria antitoxin was produced in a horse named Jim. There was a proliferation of all these companies making these anti-seria. And this turned out the horse had tetanus at the time that this was made. And this was transfused to a large number of children. I can't remember the numbers, but maybe 20 children died of tetanus from receiving this diphtheria antitoxin. And that led to the Biologics Control Act, which actually came before drug regulation. Also interestingly in terms of the NIH FDA tradition, the Biologics Control Agency was off and on part of NIH over the years, an interesting bit of history. And then a very interesting thing that David Relman had called to my attention, and I was also aware of, though, is that in the 60s and 70s, there was enthusiasm for replacing bad bacteria on the body with good bacteria. Kind of simplistic notion, but not totally unrelated to what we're talking about here today. And less virulent strains of staff were used to reek the idea of being to replace the more virulent ones and prevent serious staff disease in children. And quite a number of children, I think, was about 5 percent of those treated, then developed infections, some fatal, with the Staphylococcus that was transferred. Another point I always like to bring up is that having, this gets to that ecosystem idea, but even more individually, every part of that ecosystem is finely tuned. I mean, we're an amazing machine. So healthy biology, and especially with respect to immunity and inflammation, is a balance, a constant balance of modulating factors that allows us to be resilient when we're challenged from one end of that spectrum or another. So immunity and inflammation versus tolerance are balanced. Clotting is balanced. Yet many of our medical interventions are blunt. And so one example is one of the greatest therapeutic advances we have is TNF blockade. It's revolutionized the lives of patients with rheumatoid arthritis with inflammatory bowel disease. But yet, as most people know, it does give you a failure in your response to certain things, and you get serious brain viral infections, histoplasmosis and lymphoma. And obviously, if we clot too much or too little, we have problems. So we need to consider that altering the microbiota could have similar implications. But on the other hand, it could provide an opportunity, especially with people going into it, like all of you seem to be doing, understanding its subtleties, to get the porridge just right and not have it be too hot, too cold. Now, a little bit about observation and causation. I'll go through this quickly because I think everybody's pretty familiar with this. But just to say that looking at observational data can be very tricky, things are traditionally assessed such as the strength of the association, its reproducibility, whether there's a dose response, which may not be very relevant to the microbiome, and whether it's biologically plausible. Of course, the biggest problem is what we call confounding. That is, are there other factors that we can't measure or correct for that are responsible for the association? And confounding can itself, many people don't realize this, lead to very reproducible results. So just because somebody finds constantly that the microbiome is associated with X disease does not mean obviously it causes the disease, because if there's consistent confounding factors like diet or something else, you will still have a consistent finding. And the most extreme example that I sometimes use is, does the Starbucks in neighborhoods cause higher income in a dose response relevant way? Okay. And if there was ever an observational area subject to confounding, certainly the microbiome is it, it's very complex, everything influences it. If it influences everything else, things like socioeconomic status may influence it, although I haven't seen studies of that, but I suspect they've been done and I'd be interested in that. And the other thing I wanted to mention is I see all these wonderful dendrograms and all this incredible data, but I used to see this when people were dissecting the immune system and they had all these T cell subsets and you could, in the clotting system, you could explain or hypothesize a role for anything being up or down-regulated causing virtually any disease. And I think systems biology has both given us tremendous insights, but it's also made it possible to propose biologic plausibility for nearly anything. So we have to be a little careful of telling stories and then believing our own stories. So taken together, I think for microbiome interventions, these will make controlled clinical trials very important. Now there are some unique issues in microbiome manipulation and implications, including parallels that I really like and I've heard here to ecosystems. I wanted to mention a few other ones. In reflecting on this, that how reminiscent this is of ecosystem issues, I also realized that we at FDA, et cetera, frequently see these and in clinical medicine in other domains. For example, states of health and disease are frequently not binary and their establishment or treatment may require multi-step process. We just heard about hit and miss kind of things. Well, late in the course of a disease, an intervention which could have been effective early on may not be effective. And this goes against how we normally do clinical trials. We normally treat severely ill patients first because they're more desperate for cures. They have a higher potential benefit, but they in fact may not get the benefit. Interventions in the microbiome particularly may or may not persist. So the fact that something doesn't work may be not so much that you didn't have the right idea, but that the intervention itself doesn't accomplish that change in the biome. And then the most important outcomes, whether positive or negative, may be long term. So people are looking at chronic disease in many of these cases and obviously those are very hard studies to do. Now, I think one of the areas where the investigator and innovator community comes into contact with FDA is the products and their characterization. I wanted to say a little bit about that. These are really unique products. There's a need for defined and characterized products. As we've heard from several people, there's a need to define do we need these mixtures, these complex mixtures, can we hone them down and what are the tensions between that, between a viable ecosystem since that's what it is, and reductionism and how we approach that. I mentioned the potential for adventitious agents and I will say that sometimes one man's non-pathogen can be another man's pathogen, so we need to think about that. Genotype versus phenotype, what's necessary for the desired effect, the stability of organisms and mixtures, both as a product and in the patient. So based on these issues, the kind of information FDA typically wants to see is the source, the history, the genotype and phenotype, the antibiogram, the stability of these products. And we do take a risk-based approach to that, so for early clinical studies, you need less data when you're exposing less people, depending on what the source is, you may need more or less data. So we have guidances, there's a guidance on live bacterial products that can be helpful, but I wanted to emphasize that guidances are just that and we really encourage people to come in and talk. And if something seems unworkable or you have alternative approaches, please propose them. What about clinical trials? How do we decide whether to do them? I think this is almost like anything else. Is there strong evidence of causality? Do we have a mechanism of action that supports the proposed intervention? Is there evidence for safety and efficacy, whether in animals or humans? Important one to think about and we've seen lots of problems in other complex products like cell therapies. Can you reproduce the therapy in the future? And that gets again to characterizing the intervention or the product. If you can't, you're subjecting people to risk without benefit. And if you can't, you're also very frustrated because we've seen a lot of things that worked one time and that they couldn't be made to work again. Specific issues, I mentioned fine-tuning the intervention, product characterization, what other things. One interesting point is the challenge in population selection and this could be a benefit. If you can identify this dysbiotic subgroup that needs the intervention, you may be able to get away with much smaller clinical trials than if you just tried to treat everybody with obesity or everybody with inflammatory bowel disease. Monitoring the effects on the microbiome both to establish a dose response or an intermediate endpoint. And I think importantly for development of biomarkers, it raises the complexity and importance of appropriate sampling and data analysis plans. I also want to mention there's a unique opportunity in terms of biomarker and diagnostic development and that building good data collection and assays into studies could allow us to define biomarkers and provide disease insights. And I mentioned long-term follow-up and I know I'm tight on time so I'm going to zip along here. So what you really want to do is think through early on and I'm always surprised at how often even sometimes very sophisticated product developers collect all this data and then really decide what they want to do. I think you have to learn from your data and you could change your intention based on the data but it's a good idea to have a sense upfront what the candidate patient populations and indications would be, the criticality of ensuring the ability to reproducibly produce and characterize the relative product or intervention, the study and relevant disease models, careful consideration of safety issues, and then design stage clinical trials with relevant endpoints. And please consider sampling and assays for candidate biomarkers and doing, you know, in multi-center studies, central labs with well-validated assays can be very important in this. I think our current clinical research enterprise in the United States probably loses as much useful information as it actually collects and analyzes. I think this is emerging science that we all can and should learn from. So I'm delighted to be able to be here and encourage this field and this interaction, hopefully inform it a little. We know we are ourselves complex ecosystems. We have both vulnerabilities and resilience. We don't want to fix the resilience. We want to fix the vulnerabilities. I think the promise of recognizing this reality reminds me of the recognition of our broader global ecosystem and environment and the need to support its balance and offers great potential benefits. For all these reasons, it should be based on the best science and a healthy respect for the complexity of nature. And personally, and I think in general, we're optimistic about the impact of microbiome science and welcome a really constructive engagement going forward. Thanks very much. So we have time for some question. Hi, thank you. That was terrific talk. I'm with the Office of Dietary Supplements, Cathy Camp. And I wondered if you could address some of the issues that we face with things such as probiotics that are actually regulated as dietary supplements versus drugs and their use as targets to the microbiome, particularly with infant formulas. Thank you. Thanks. Well, I don't have particular expertise in the area, but I think as one of the speakers had said on the first day, and you're probably an expert on this, FDA has somewhat limited oversight over dietary supplements, except when something goes wrong. So manufacturers are required to support, report serious adverse events, and FDA will investigate those. And I think it's safe to say that there haven't been a lot of serious adverse events from the commercially made probiotics. That said, if they, my understanding is if they use ingredients developed before something like 1994, they get no actual pre-marketing review at FDA. So essentially they can be out there selling these probiotics without an unbiased review of the data. So I think it's a little bit of a buyer-beware situation, but I'm not aware of a lot of harm. The flip side of this is there may be opportunity. We've seen that some of the clinical trials have been done with some of these commercially available preparations, some of which have shown some interesting results, for example, in infant diarrhea. I think one of the things I've heard a little, we worked with NCAM on this issue a few years ago at NIH, is it would be great for these companies to bring forward more information about their products to engage the research community. Some companies have been very willing to provide researchers with information on the manufacturing and content of their products, but in many cases researchers have a hard time knowing what an actual product is. I will say not to impugn any specific product, but in some cases when products have been looked at it's been found some of these do, but others don't contain live organisms etc. So I'll leave it at that. Maybe I have a quick question. One of the issues is a lot of the whatever you want to call them probiotic or other different product that are derived for the microbiome is that they don't really fit one category. They can be devices, they can be food, they can be drugs. Is there any plan at FDA to kind of centralize maybe a an entry point for us who are developing or thinking about those things to guide us through the complex FDA structure? I think that's a really good question because it is a complicated system and I want to point out it's that way not because like we like it, but that's because the laws are different in those areas and companies decide how they want to submit their products. I think in general your entry point is the center for biologics because what we're talking about is using a microorganism or an intervention and if that's wrong they can help you elsewhere but for a medical indication to treat, diagnose or prevent disease and that becomes a medical intervention and does engage a pre-market review process. In some cases that might not be the case or for example a trial of an antibiotic to alter the microbiome might be with the drug center but we I'll bring that idea back and we can consider you know certainly we can look at our web interface for example and try to improve or try to put up a page that might make it easier to enter and navigate the system and the other thing I would offer is you know people can contact my office too because I mean there's always this this concept of which you bring up which is curing disease but what about maintaining health what does that fit? Well the you know I think maintaining health in terms of objective measures i.e for example one of the things you might want to do is have a resilient microbiome so that you prevent diarrheal disease or prevent atopic dermatitis that still comes under prevention of disease and in general we would encourage that kind of engagement. There are these things called structure function claims so when you and that's what the dietary supplements use and that those are those sort of interesting things you'll see in advertisements like this will help your immune system or this will help your joints and again that's related to the what the law is and our staff do their best to sort of keep an eye on that but you can see there's a very I think for credible scientific purposes in general you're talking about preventing treating or diagnosing specific illnesses. One last question and we'll go to Greg. I'm Gilman Gray from the Child Health Institute thanks very much. I enjoyed your overview. We're interested in the oligosaccharides in human milk that have antibacterial or antiviral activity especially two fucacil lactose. Do you see any future in honing down on specific oligosaccharides that may have biological activity and adding them to formula or giving them in a pill? Are we are we going down the wrong road here because we're kind of interested in doing that? Yeah well I don't feel like I know the science there enough to comment on it but what I would say is as a concept I think that if it is scientifically sound that is a really good direction to go in other words if we recognize simple interventions that can accomplish or aid or a bet you know what we see the microbiome as a whole functionally engaged in those are a really good idea. I could you know again if I don't know if you've been in contact with folks but if you sent me an email I could try to get you connected to people with expertise in that specific area. There are in the food area there's an issue where if a company can file a petition with the FDA for what's called generally recognized as safe to sure if they can show based on scientific literature etc experiments that a certain substance is safe that actually is often if that's approved by the FDA it can be allowed as an ingredient but obviously that doesn't answer the question then of is this effective in what you're trying to do which is you know prevent or ameliorate disease in children. Yeah so ours is really a medical model and we would have to go through the FDA and get all kinds of clearances to pre clinical testing and all of that right and we need gram or kilogram quantities of it right now there are only most exist in microgram quantities. Yeah what I what I can say is it is hard to develop drugs particularly if there's not commercial interest behind it but if there's promising information you know NIH may have ability to help if you have business partners they may have ability to help. There's tremendous we've seen tremendous power in the patient advocacy community to bring together groups who can try to help move products forward and one of the models we're trying to put forward at FDA is to try to help you with that so you know if to try to help lay out a pathway help identify what the issues are rather than just have you come in and feel like you're having to meet you know commercial requirements now the flip side is there are safety issues you know many active pharmaceutical ingredients come from all over the world there's a lot of issues about quality and safety of the supply chain so those are important things to consider too. Great I'll take you up on your offer you'll be hearing from me thanks so much. No you're very welcome. Well let's thanks Dr. Goodman for a great talk. I'd like to welcome everybody to the next session this is the first of the translation sections and this one is on the body microbiome access I'm Carl Baker I'm program director for or for keratinocyte biology and diseases in the division of skin and rheumatic diseases of NIAMS and I'm going to be chairing the session this morning. I'd like we have a very full session so I just like to remind speakers to stay on schedule so that we can have a little bit of time for questions. So the first speaker this morning is Ted Dynan who is from the University College Cork in Ireland and he's going to talk about the microbiome brain and behavior.