Januari 11, 2017
our first speaker today is neil caporaso. he got his undergraduate degree and then a masters degree and then a medical doctor degree from rutgers in 1980. so he did a residency in internal medicine and in 1983 saw the light and came to nci
and now he has to move his office every other year. [ laughs ] just like the rest of us. so, in 2011, he became chief of the geb, genetic epidemiology branch and he is in the division of cancer epidemiology at the nci and he's going to talk to us
about epidemiology translational research and clinical oncology. take it away neal. >> thank you so much. so i'm going to give you folks a broad overview of epidemiology and a little bit about the basics. but i'm also going to touch on
some of the directions that epidemiology is mutating into. i have a fair number of slides and i'm not going to go into depth in all of them, you'll be happy to know but i'll try and emphasize the ones that are most entertaining. so i'm going to touch on some
introductory concepts, some tools, some things that epidemiologists have accomplished and what the challenges are and where we are going. so, a few foundations. let me start off by saying that our group is located in nih in
the national cancer institute and the we are part of the intramural program. as those of you are in the clinical center and in ncc r and we are part of the division of cancer epdemiology and genetics and right now i'm chief of the genetic epidemiology branch.
we have a number of other branches that work on the environmental sites. so there are a number of branches that investigate genetic causes of cancer and a number of others that investigate environmental causes of cancer.
and of course our division is generally dedicated to the ideology of canser or causes of cancer and that's the general theme of epidemiology. our studies had a lot of impact on regulatory changes and our water and gasoline, diesel exhaust, chemicals, we have
specific studies in each of these areas. and every one of these studies has had a fairly dramatic -ue can spend a whole hour talking about any one of these studies. so the diesel study for example, actually was investigated a number of times because there
are important chemical interests in this country that really don't want to have anybody regulating the use of diesel exhaust. so the head of the occupational studies brach was called to testify before congress 4 separate times to defend that
study. so, epidemiologists have to condemned with a political dimension for their work. say for farming, it seems like such an iknackuous thing, safer farming. there is a study, a giant cohort study called the agricultural
health study that investigates pesticides and herbicides. and that study conflicts with gigantic corporate interests like monsanto, that produce agricultural chemicals and don't want them to be particularly regulated. kimember can be wild of everof
kinds of conflicts. we investigate cancer susceptibility syndromes, second cancers among cancer survivors, impacting the medical of radiation et cetera. a lot during the la last year, a large study has been launched to investigate whether doses of
vaccine is as effective as three doses. easier to implement man two or three. there are collaborations all around the world. and you can go to our website and find out all the kinds of studies being done if any are
interested in epidemiology as a career path. you can learn about hundreds of different studies in different areas that are ongoing. and there is risk assessment tools and risk assessment is a big deal, particularly if you want to screen for cancer.
one of the discoveries that has coming out of our group over the last few years is a quantitation of the idea that if you're going to screen for cancer, say lung cancer, to do that efficiently, you really want to identify the people at highest risk for that screening.
it seems like common sense but to show this series of papers including in new england journal that actually showed this rigorously and mathematically, that you want to generate a risk model and then show what that risk model that will you're going after the people at
highest risk to do your screening efficiently. how do you do that with lung cancer for example? well, you want to identify a characteristic of smoking, for example, that are most associated with lung cancer and then you want to know what other
features are also most associated with risk of getting lung cancer? things like do you have copd? how old are you? your family history of lung cancer and a factor our group identified which is called, time to first cigarette.
so if the amount of time that between when you get up and when you smoke your first cigarette f it's less than five minutes, your risk of lung cancer after taking everything else into account, is 3-4-fold higher than someone that waits over an hour for their first cigarette.
so all of those things need to be incorporated in an efficient risk model and that makes screening efficient. and although the national lung cancer screening tile showed that there is a 20% reduction in mortality with helical cp screening, the economics of
introducing ct screening is really difficult and expensive and you're not going to do it unless you have a better risk model. so, some introductory ideas. epidemiology refers to populations and disease and populations as opposed to
medicine, which is disease and individuals. and it's an observational science as opposed to an experimental science. the idea being that we might look at a population and identify some individuals who are smokers and some individuals
who are non-smokers however, we don't give cigarettes to people and say we want you to smoke those and we want you to not smoke. that would not be ethically permissible. so, that is one of the distinctions between an
observational and an and it's a particular method logic weakness that those who don't like the particular conclusions that epidemiologic studies come to often criticize. and those are usually individuals that have an interest in the outcome, for
example, tobacco companies. and so when tobacco companies criticized the first case control studies of cigarettes, they said, it's an observational study and so the results are not reliable and we don't believe them and you all should just keep on smoking.
so, this is really the point i just made that epidemiologists are ethically prohibited from doing experiments on people. and this weakness was exploited by the tobacco companies. subsequently, epidemiologists are concerned with the idea of establishing causality.
what our study designs allow us to do is formulate the statistical associations in those associations can be quite strong but they may not be mechanistically compelling. so we like to assemble evidence from other domains to make those arguments as compelling as
possible and i'll give you some examples. so the goals of epidemiology generally identify the causes of cancer. quantify the risks and elucidate the mechanisms where possible. epidemiologists are always traditionally been concerned
with public health. so you'll hear the statement all the time that the people who cleaned up the drinking water saved more lives than all the physicians in blah, blah, blah. that kind of thing. but epidemiologists really do have a public health orientation
and take this very, very seriously. epidemiologists emphasize prevention. so lots of examples, all sorts of interventions that prevent disease, there are a number of downsized to prevention i have to be frank with you about.
one thing about provision is it takes time and very expensive to demonstrate a preventive intervention is effective. and it's not very dramatic. you kept really see the disease that you prevented. you don't have a grateful person that comes to you and says,
thank you for saving my child. they don't know the vac seep saved their child. they take it for granted. and if their child 10 years later develops a rash, maybe it was your vaccine that did it. so not always so grateful. dramatic treatments are
effective with politicians. whereas preventive interventions tend to be more difficult. and so, it really takes a certain amount of i don't know, intelligence and real orientation towards public health to be dedicated to this. and also you have to be willing
to fight political opposition. so whereas years ago, we had tobacco companies and that's a little bit pass a, now you have companies that pretty much advocated processed foods and foods that have no nutritional value but basically are pure sugar and are marketed to kids
and contribute to the obesity epidemic. and so, it's very clear that obesity is strongly related to diabetes, heart disease, and about 8 different cancers. so it's going to take probably a generation to develop that public health consciousness and
have kind of intervention that is will be effective and meantime, all of those companies are fighting tooth and nail to prevent that from happening. and the same thing with polluters and companies that make chemicals. so you really are in a political
battle. and finally nobel prizes. in spite of the fact that the discovery that tobacco was the cause of lung cancer and tobacco in the next century will likely cause a billion, that is a billion with a b. excess deaths.
so what was that number around in your mind of the a billion with a b. that's a lot of deaths. that's more than the number of deaths than a lot of other causes. nevertheless, the discoveries and -- we can go through a lot
of discoveries, we could have a whole talk on just the series of discoveries about tobacco. no, no bell prizes for that one. for some reason, the scientific plus the political fight to do that was not considered as important as say ublowing in the wind or i don't know -- that's a
joke. whatever the latest chemist did to win the nobel prize. or the latest medical nobel prize. so, what are some of the concerns? son bias, which is a systematic deviation from truth.
and an example of a kind of bias is participation rates. so, epidemiologists are very worried about participation rates because if you have a low participation rate in a study, the subjects that are in the study are not going to be very representative of the general
public and generally you want the people in your study to be able to generalize the results from your study to the wide world out there. if you can't do that, your results are of limited value. so that is an important thing. and in fact, if you take a
careful look at many, many studies, the results aren't so generalizable. so let me give you an example. a u.k. biobank. now if i said to you, is that a representative study? you think about it, wow the general population of the united
kingdom and it's associated with their health care system. yes, that sound like it would be pretty representative. in fact, less than 15% of the people that are insighted to participate in that study actually do participate and you have to wonder why is that?
maybe they don't want to give blood. maybe they don't want to participate in a questionnaire that takes an hour. maybe they just don't want to be bothered. maybe they have been drinking a lot and they are like, i have a
headache today. i don't want you to know how much i drink. or whatever the reason is, they may not be so representative of the general population and whatever the reason is, it detracts from the generalizability of that study.
so, compromises always have to be made but it is something you need to keep in mind. so, i'll give you one example from real life. in 2003, we did about the time the largest lung cancer study in the world. and we had site visitors that
came and they said, look, if you can't get a high participation rate, we are not going to give you the 8 million dollars you need to do this study. and so, we said okay, we'll do a pilot study to get a high partiseration. and the first thing we did was a
phone survey. we called up people and said, have we got a study for you! you get to give blood and do a questionnaire and it's really great study. what do you say? 30% of the people said they would participate and that
wasn't high enough. we then sent them an invitation letter and followed up by phone. we said if you're in the hospital it's okay. we put advertisements in the hospitals and magazines. we offered them a cash award in gasoline which in italy at the
time was a big deal. we got their physician to write a letter and said we would come to their home. that got it up to 49%. still no good. we couldn't have done the study. so then, we got new interviewers.
so for the men, we got attractive nurses and for the women we had male nurses. we had the physicians call. we give them gas coupons. i said that already, instead of money iy had ads on t.v. by a very popular night show host. we had a better invitation
letter that was written at a fourth grade level instead of a more sophisticated level. we got a letter from the mayor, and we had a 1-800 line. with those innovations, we got a 73% participation rate which was adequate. there was really hard to do.
that pilot study over a year to go about a million dollars. so, it's not easy to do a study to get this kind of -- something to keep in mind. epidemiologists worry about the controls in your study. population controls are really expensive.
what is population control? population control is representative of the population and the issue with population controls is that it's impossible to get them these days. in the past, you used to use rvd, which is random digit dialing.
now if you do random digit dialing, if you have a land line, you get a nuisance call, what do you do? you don't even look. you hang up. it's like don't bother me. so, it's incredibly hard to get good controls.
and so you have to make compromises. it's a really difficult thing. and then when you read about high response rates in papers, you say what was their response rate? was their response rate really 70%?
you need to be very critical when you believe them because it's very hard to get -- it was hard 10 years ago in italy. it's even harder to get or to do that now. and i don't have time to tell you about all of the theoretical advantages of good controls.
but there are advantages. if you call an epidemiologist because you have a study, and you like some help, you have -- the epidemiologist will ask you about your study design, where your controls came from. do you have covariatessan? issue i get -- covariates.
-- an issue i get all the time is somebody applies to use the plco or the nlst file specimens and they say, i have a fabulous marker for lung cancer. could i please have some serum from the national lung cancer screening trial because i'm going to test it versus your
controls and i know that my marker is spectacular to detect lung cancer; because we tested it in some lung cancers already and it picked up the lung cancers. so my first question is, did you adjust for confounders? they say, what is a confounder?
i say well, did you match them on smoking? because you know what? most people with lung cancer and if you just took lung cancer patients that are smokers and you took controls that are non-smokers, your wonderful biomarker may be a really great
biomarker of smoking. it may be coating. it's a great lung cancer marker but of no use whatsoever as a lung cancer screenings because we know already we can ask people if they are smokers. we already have that in the risk so we don't need it.
you need all of the covariates. so, that's an example of something you need. so bias, confounding, original hypotheses meaning you -- there is a dinners between -- here is another one that you get a call from someone who says, i tested a biomarker and it predicts
liver cancer in a cohort. it's strong associated with liver cancer. and what you find out is that they tested the biomarker versus 30 endpoints. and there was a statistical association with one cancer. and so, well, what is the
probability you get air significant association at a p value of .05 for 1-20? you test at 30 endpoints and of course you're goi to find one that is significant. so, that's called date at dredging. and power calculation.
some other issues are very important. you see, she out lived all her doctors, she drank every day. she had bacon every day. how do you sprain that? epidemiology is a probabilistic science not deterministic. you always get individual
outliers. individual outliers are a mathematical certainty. you're always going to get them. so it's just something to keep in mind. one very cool tool that is still being used are maps. so what thesis tend to show you
is that cancers geographically associate. here is an example showing the red in the south and this is incident, mortality from melanoma. and obviously, this is due to sun exposure. and melanoma is associated with
sun. so this is a pattern indicative of sun. lately we have been doing some studies, circadian variation and looking at top variation and capser rates by time zone. so we look at the time zones and we can see associations with
time zones. this is the whole emerging area. and a lot of studies are done with surveillance epidemiology and end results program. so, this covers 26% of the u.s. population. i believe it is 11 states and has data on incidents and
survival and patient demographics and tumor morphology and histology and a lot of other cool information. and so, it is ecologic data-mining you don't have individual risk factor data that covariate data i told that you is important.
but it is very, very valuable and important and it is searchable. there is a tong of information that you can analyze line. and an example of how interesting this data is, if you look at cancer incidents data, you see a rise in cancer in
1994. so does anybody know what happens here? this rise in men and not a corresponding rise in women. so, this was when they introduced psa screening. so, the rise was due to prostate and discovered a lot of cases of
prostate cancer due to psa so this was actually a glitch in the data and inaccuracy or a -- not exactly inaccurate. the data was accurate but it was a -- something that had to be taken into account and more prominent in african-americans than whites.
there is incidents and mortality data and many times when you see a big difference between incidents and mortality, one of the things it tells you is that there is an effective treatment and in fact, there is effective treatment for many kinds of pediatric malignancies as you
know. so, i mentioned to you that because epidemiology depends on statistical associations, there is a lot of attention given to proving causation. and so, there are classical criteria for proving a cause. and some of these criteria are
that an association that we believe is reliable as a causal one, should be a high risk, a high relative risk t should be consistent over numerous studies. it should show a dose response when you have more of whatever it is, the risk should go up.
the cause should occur temporally before the cancer that is always important. and the biology should make sense and should be some kind of a mechanism that is in there. lately there are other and more sophisticated and more direct approaches to get at this.
in basically involves looking and doing a genome-wide association study on the factor and on the disease and looking to see that they share genetic ediologgic factors and that would be an element tending to support a causal association. and molecular epidemiology
involves using biomarkers to infer mechanism and mediation analysis involves a more complex kind of causal analysis. so epidemiologists use all of these to try and infer causality and a more sophisticated way. this shows what cigarettes, how three cohort studies showed a
beautiful dose response, a consistent dose response with cigarette smoke. and it mechanistically made sense and this is bios car, a pathologist in louisiana who did famous studies with smoking beagles where he attached smoking machines to these
beagles and then dissected their respiratory tree and showed that the exact same progressive changes occurred in the respiratory tree as they developed preaplasia. and some beautiful work also in cohorts that showed when people quit smoking over the years,
their rates of lung cancer slowly and consistently fell to approach that of nonsmokers although in never quite equals that. so it approaches a relative risk of one but never gets there. i'm overwhelmed with the idea of going over what epidemiologists
have accomplished but. i'll tell you that looks like coffee is not associated with any particular cancers or adverse effects. a nice study by my friend neal friedman. there has been a lot of studies, breast implants, consider noble.
oral mouth wash, a lot of studies have been null. cell phones are not strongly associated with brain tumors although i understand there is new data coming out which may show some small associations but this is probably an area where the radiation epidemiology group
gets the most calls. people are most worried about cell phones. the big worry was for cell phones over 10 years ago when they had a lot more emissions than they do today. in general, modern epidemiology identified the general and
specific causes of cancer and advocates of public health, the identification of tobacco as the major causal factor for lung cancer and seven other major tumors all which are difficult to treat and the role of secondary tobacco smoke. so, i would just spend a second
on secondary tobacco smoke. you realize that the discovery that secondary tobacco smoke was associated with a number of of capsers and other conditions is what allowed clean air legislation to go forward and it means when you get on an airplane or go in a movie
theater or go to a restaurant, you don't have to breathe in other people's smoke. and so you're saved from not only that risk but the announce of being exposed to all of that. and that also contributed to a social -- that makes smoking less acceptable and in turn
makes the per capita rate of smoking less -- go down. this is an enormous public service that came from those epidemiologic studies. okay. the general respect is the for cancer are age, as we like to say, e and g and combinations of
e and g. tobacco still the biggie. diet is going down in importance and there is a -- i would say a crisis in nutritional epidemiology. and it is worth taking 30 seconds to tell you about this in that most of the studies of
dietary factors purported to be either protective or risk factors for cancer have proven to be challenging and difficult. the hope now is by using molecular epidemiology techniques, biomarkers like techniques like metabolomics, to figure out a mechanistic basis
to understand some of the associations. so there is data and for instance ir data that still supports a weak association of meat consumption with some cancers and a weak protective affect of fruit and vegetables on a number of cancers.
but most of the associations with individual nutrients like b vitamins, have washed out and don't seem to be present. it's been really tough to prove nutritional associations with and that said, most cancer is due to the environment and evidence for this is that there
are dramatic differences in rates of different cancers in different countries. so, melanoma between the highest, australia and the lowest, japan, the ratio is over 100. now some of this is due to genetics.
so a component is due to genetics but we know from migration studies where people of one group go to another country and typically when that happens, they acquire the rate fairly rapidly of the new country. so this occurred with most
cancers and i'll tell you one or two exceptions. for instance, best cancer and prostate cancer. within a generation, if asians migrate to san francisco, they acquired the rates of cancer of the united states within a few decades.
now one or two examples where that is not the case. for instance chronic limpcytic leukemia. but that is the most common adult leukemia, has no known environmental causes but it has some strong genetic causes. so that may can an exception.
here is an example of an environmental cause. when we look at the maps, a province in china, and the county specific mortality rates based on the map or super high, when you go there, what do united states see? they have indoor ovens that
create incredible levels of indoor pollutants so they have ovens you should the beds called tangs, that they use for heating. and those smokey coal ovens result in extremely high rates of lung cancer in that province. and the cancer maps also showed
a high rate in montana and i had a map here. that slide somehow dropped out but there was a copper shelter that contaminated the with ars inic and that accounted for a high rate of lung cancer around this shelter in montana and they removed the shelter, the rates
went down. so, this is just a reminder that tobacco is a bad guy and i think i told you a little bit. i'm not going to go on and on there. but one good thing about tobacco in the united states is that after the surgeon general report
and all of the public health consequences, rates of smoking in the united states have steadily declined and that's a combination of individual knowledge and public health efforts and the social efforts. so you can see these rates continue to fall and the rate of
adult smokers in the united states is now below 20%. still rather shockingly high when you think that still something just under 1-5 u.s. adults still smokes. worldwide anticipating a billion deaths in the next century from smoking and most of that is
because the tobacco companies have exported their industry to developing countries like craned and china. and i told you a little bit about environmental smoking so i won't dwell on this too much. alcohol is the number 2 carcinogen and it's associated
with a number of different five top ones are oral, favorrics, esophagus, larynx and liver and interactions with smoking. and ionizing radiation is also a big extrinsic cause. i'm going to move ahead to say a worded about some newer studies
and skip over some of these because i think this is material you can go over and look up more on your own if you're interested. chernobyl. non-ionizing radiation is a major cause of skin cancer. and aging of the skin and the
microbiome is an extremely important area. hpv over the last year, investigator in our group sequenced the hpv virus and that is an extraordinary finding and discovery and what it is showing is that some characteristics of the virus contribute to
carcinogenissity, so you'll be hearing more about that in the literature. and this is prevalent in colorectal carcinoma and i think you're going to see a lot of discoveries with regard to colorectal carcinoma as a number of studies in the feco
microbiome mature. and a lot in occupational and i mention the diesel exhaust so what are some challenges? so on the environmental side, for a number of cancers, we still know nothing about risk factors. so we don't understand much
about the risk factors for examples, chronic limpcytic also, beyond a few risk factors for breast factor, hormonal factors and alcohol, we don't really understand the risk factors for many common cancers. so i mentioned breast but really prostate is a better example.
age and family history. not so much is known about the risk fact another for prostate how genes and environment work together is not well understood. and many of the potential causes are poorly studied. i mentioned chronic and lymphocytic leukemia.
some of the channels in i don't have time to go into a lot of the issues with diet but that could be a whole talk. in our study in italy a told you about earlier, we did look at fresh red and processed meat intake and did show a dose response and possibly the most
consistent finding in nutritional epidemiology is that the more red and processed meat you eat over a variety of cancers, the more your risk increases. so treating meat as a conned cement probably not a bad idea. so genetics?
yes and no. for common cancers, gwas studies show numerous genes associated with virtually every cancer however the risks are small and a big issue is that when you put all of those small risks, by small i mean a relative risk of 1.1 to 1.3.
so when you put a risk factor of 1.3 and a risk model and take 3-5 of them and put them in there, what happens? unfortunately not too much t doesn't increase the risk a great deal. so, it hasn't proven extraordinarily helpful.
has it taught us a lot about the mechanism? a little bit but not so much. so we still have a ways to go on the genetic side as well. however, one thing that is well established is that virtually all cancer is associated with genetic changes in the tumor and
in fact, our group, our division is establishing a new group to characterize those genetic changes in tumor. one problem is that to do a good job in sequencing tumor, you really need fresh tumor. and fresh tumor is not always available.
unless your study liquid nitrogen and has it in the operating room when the fresh tumor is obtained, sequencing can be difficult because it's hard to do it from sections. so this is something that the next generation of studies will have to be established to
collect. the other thing you need if you're going to do this is very good covariate data. so, there are fresh tissue but many times they don't have the covariate data you really want that good exposure data in addition to the good clinical
data and the high quality tissue to do these kind of studies. okay, so i'm not going to go into the details of genetic studies but there are studies that focus on germline and studies that focus on tumor tissue and there are families of population studies and studies
that look at everything, diagnostic and studies that look at specific candidate genes. virtually every tumor has a genetic component and this is shown by one example in lung cancer every study when you look carefully has family history and tends to show excess risk
associated with family members with cancer. and here is an example. 1.57, or 60% increase risk if you have a first degree relative with cancer after adjusting for every other risk factor. so, i'm going to skip over some of the genetics and just tell
you, touch on two more topics. what is molecular epidemiology? traditional epidemiology started out just by relating exposures to disease. so how many people smoked? how many cases and controls got we do some simple calculations and derm if there is an
association. molecular epidemiology includes an assessment of genes, users biomarkers to assess internal dose, early biological affects, intermediate markers that assess allotteddered structure or function and maybe tries to get early disease.
and takes all of these markers into account. so this is the idea of molecular epidemiology which emerged two decades ago. aise said, we did a lung cancer case control study where we actually got the tumor and put it in liquid nitrogen and gave
it to the pathologist and section a piece of tumor and put to the immediately in liquid nitrogen. and we carried this with a questionnaire that gathered that covariate data and so for example, when we looked at the information on meat, we would
assess things like doneness because there was data that suggested that the amount of heterocyclic amines in meat was related to the time and temp that you cooked meat. so burnt meat had a lot more heterocyclicness. so it would address this
question. and we would assess how people would respond on the questionnaire to how they cook their meat by pointing at a picture. so molecular epidemiology accomplished a lot of things. examples showed hpv was the
cause of 100% of cervical cancer before they actually were able to assess hpv. they would get associations with, for example, number of sexual partners. so that was a good piece of information that suggested an infectious agent could be
working here and sequencing the virus establied this in a much better way. another study showed that cutting down smoking is ineffective. why? because biomarkers studies show that you alter the topography of
how you smoke in a way that you still absorb the same amount of carcinogens because you're addicted. you're addicted to nicotine. so your body all thers the way you inhale significant relate smoke so you still get the same amount of car sin yens and of
course gwas studies were all baseline biomarkers. and then is there integrative epidemiology that adds behavior like the time to first cigarette and the outcome. so not only do we look at disease but we look at whether people survive.
so, if we -- that is the term integrative epidemiology. it's added on to molecular when you add behavior and outcomes. and a lot of these studies are done in consortia. and consortia tried to harmonize the different questionnaires by
using online resources. i have already used my hour so i'm going to take two more minutes and then stop and let you ask a few questions. just to tell you that there is a lot of exposure areas that have not been able to be studied with traditional approaches.
and just going to mention these, these ar the future. so, one is sleep. and traditional questionnaires tended to ignore sleep but sleep is incredibly important and the new generation of apps that many of you probably have like the fit bits, automatically record
sleep. and it provides a wealth of data and length of sleep and timing of sleep is clearly associated with the number of conditions. so this is an area being explored. physical activity and inactivity is also super important and
there is a recent study that are showed associations with cancer. very difficult to collect information in a questionnaire. but again, these fit bits and other guyses can collect information. heart rate both resting heart rate and variability neheart
rate are important. and need to be studied. social factors there has been data that from framingham showing that a lot of important traits like obesity and smoking tend to be linked to your social network. studies showing that most voters
stepped to only know voters who vote same as they do. so, this is a factor that needs to be studied a lot more. your geographic location, your zip code, is an incredibly important determinant of what diseases you ultimately suffer from.
including your economics. smoking can be determined by these devices, by the movement of your ample just as your fit bit records your steps based on a certain movement. smokers can be characterized by the same kinds of movements. whether climate hasn't been much
studied but can be now and circadian variation is an area i'm spending almost half my time studying and all i'll say is that circadian dysregulation likes like it has very important associations with poor conditions, mood disorders, obesity, diabetes and cancer and
i don't have enough time to go into it now. i'll stop there. thank you. [ applause ] >> off mike mikes. >> so, the use of drugs in smoking cessation is incredibly important because the point
prevalence of success at smoking cessation at one year is only about 5%. in other words, someone says i'm quitting. i have quit. i stopped smoking today and i have quit. one year later if you come back
to that person, 5% will have succeeded. however, if you -- in succession, apply a number of methods, the first is counseling. simple counseling at least doubles the rate of success. then nicotine replacement.
again, doubles the rate. so you are going to get up to 15, 20% by nicotine replacement. then, first generation drugs that also help a lot. the best of the drugs isgenics, which actually goes to the nick 10ic receptor and is very effective.
now the problem with chantics is it had one of those black boxes because apparently people had a variety of nasty side effects which may have been related to genetic subtypes, and i believe and i'm not up on the very latest, the fda removed some of those restrictions recently.
but chantics is a pretty effective intervention and the combination of counseling and multimodality drug use can improve smoking cessation success to the range of 50%. so it's clearly effective and i would say that anyone who is contemplating smoking cessation
should definitely get counseling and try one or more of these methods to maximize their success and also if they tried in the past and haven't succeeded, that doesn't hurt them at all. each success of chance improves the chance they had will
succeed. so it's tremendous health benefits of quitting smoking at any age. [ off mic ] >> so, it depends on the kind of outlier and statisticians have a variety of techniques for dealing with outliers.
i guess for epidemiologists, the most common technique is grouping. so, you may have someone with an incredibly high dose of something and you just group that person. >> yes. look at risk by quartiles.
so, there is a number of approaches to deal with you also worry about the quality of the data soy not usually a big problem. >> that is a great question and i think the key items are that a slide i showed earlier on but the quality of the control
group, the size of the study, and the confidences intervals, it's really important. don't want to just look at the p value. but the confidence interval. so a statistically significant result should not include one. so you might say an odds ratio
of 5. but if the odds ratio is 5 and the ratios are 4.5-5.5, that's a very strong result. if the confidence interval is 1.1-20, you really want to look at that study carefully. i think the main thing is read the methods carefully.
and be suspicious. look and see if they quote a response rate. how many people actually participated in the study compared to the number that were invited? and then the other thing is covariates.
what did they adjust for? i feel bad for people looking at these studies that don't know the details because many times i see, readjusted for smoking. i look at how did you adjust? we adjusted by saying they were a smoker or nonsmoker. what?
what that means is you lumped -- how did you treat former smokers they are at risk? so you took current and former and nonsmokers. did you adjust each one as an indicator variable and the strongest variable related with smoking-related disease is the
duration of smoking. so if you didn't take duration into account, then most of the affect of smoking is still there even after an adjustment for current smoking. so you have to think wow, smoking is still the most important thing.
so if you did a biomarker study and just adjusted for current smoking, still duration and quantity and intensity of smoking are the main drivers and you never know that f they said we adjusted for smoking. sorry. so i put my information there
and you have my handout. any questions, don't hesitate to e-mail me. thanks a lot. >> we have one announcement. we are going to be doing pathology core visit right after this next lecture. so people can just wait out in
the hall and then i'll walk you over to building 10. i don't want you to get lost. our next speaker is haobin chen. degree from medical university in china and a ph.d. degree from new york university subsequently went to research assistant professor at new york and did a
medical residency in brooklyn and came to nci in 2013 as the clinical fellow in thoracic and gi oncology branch and now an assistant clinical investigator. he is going to tell us about small cell lung cancer. thank you to the previous lecturer.
it helped to let you know that smoking is actually the most important environmental ideology of the lung cancers but today i'm going to talk about small cell lung cancer. so here is background information about myself. terry mentioned that actually i
have medical education in china and then came to the united states for ph.d. education. and i always wanted to be physician scientist so that's why i kind of came back to do the residency training and then fellowship training. so about three years ago, i did
exact same thing as you're doing right now by attending clinical courses and i learned a lot and i feel really honored to be here to give a talk on the small cell lung cancer. so let's just dive in. so this is outline of my talk today.
the lung cancer can be roughly divided into two major types, small cell lung cancer and non-small cell lung cancer. the later also includes three histological subtype including the adenocarcinoma, squamous cell carcinoma and large cell carcinoma.
and doctor schwaubo is going to give a talk on the non-small cell lung cancer in next few months. the reason to classify lung cancer into small cell and non-small cells because small cells behave differently. small cell tends to metastasize
early in its course. and the prognosis of the patient with small cell is also much worse compared to non-small cell patients. and small cell lung cancer comprise of 10-15% of all lung cancer cases. and if there is smokers disease.
so the patient with the small cell lung cancer -- almost always have like decades of like tobacco smoking history. however, in recent years, the patient with adenocarcinoma can also develop small cell lung cancer after egfr tyrosine inhibitor treatment and these
patients are usually nonsmokers. and the small cell lung cancer is a neuroendocrine tumor. so one unique feature of small cell lung cancer it can be associated with the peritoneal plastic endocrinopathy such as curbings syndrome due to the excessive production of acth by
the small cell lung cancer cells. so morphology. size of tumor cells are small. it's also known as oval cell carcinoma and it looks like oat grain and it is small oval cell with scanty cytoplasm. so on the pathology slides here,
and the arrow points to the small cell lung cancer cell and the arrowhead points to the lymphocytes. so as we know that lymphocytes are very small cells. so, in this case, the small cell lung cancer cells are just a little bit bigger than the
small cell lung cancer cells are very sensitive to chemotherapy and radiation therapy. however, the problem is that they develop resistance very quickly. so, in the 1950s, researchers found that the compound can shrink small cell lung cancers
and later the cyclophosphamide was also found to be effective in treating small cell lung in the early 1970s, several randomized clinical trials proved that the combination chemotherapy can extend patient survival compared to single agent.
in the 1980s and early 1990s, the platinum and etoposide established role of first line of therapy and unfortunately in the past 30 plus years, it remains to be the first line chemotherapy. this is not because of lack of trying.
just that we haven't been able to find a better chemo rental min than platinum and etoposide. and the second line agent approved by the fda is topoteak an. i should have a question. does anyone know what is the mechanism of action of etoposide
and topoteak an? in small cell lung cancer, the topeo -- higher compared to the normal lung tissues. so, the small cell lung cancer can be classified into two stages. limited disease stage and extent of disease stage.
so the difference is whether the disease can be covered by a single radiation port or not. so, if it is limited to just one hemithorax and be covered by single radio therapy port, it's a limited disease. so the treatment should be radiation plus systemic
however, if the disease is more than the single radiation port then it extends disease and the treatment should be systemic so, if a patient with extensive stage of small cell lung cancer has response after first line chemotherapy then we stepped to give a patient prophylactic
cranial eradiation. and the reason is that the small cell lung cancer can spread to many different parts of the body, however, brain is -- chemotherapy so the chemotherapy agent cannot reach high concentration in the brain. so, when a patient has
reoccurrence and the brain is a common site. so in the 1990s, the metanalysis showed that the pci provo lactic cranial eradiation can improve the overall survival of the small cell lung cancer patients with complete remission with initial therapy.
in 2007, a new england journal paper demonstrated that by randomized clinical trial, the prophylactic cranial eradiation can decrease brain metastasis and improve progression-free survival in overall survival in the patients with extensive disease small cell lung cancer
and with response after initial and on the right-hand side, showed the kaplan mire curve of overall survival in patients with or without the prophylactic so, the doted curve shows the patients with the pci and the line here shows the patient without pci.
so in 2012, the u.s. congress passed the cancer research act. so in this act, it requires the nci director within 18 months to develop a scientific research framework for the cancers with a five-year survival rate of less than 50%. and are add a sense of urgency,
the act also requires the nci director to identify two or more cancers with relative survival rate of less than 20% and also has abual death over 30,000 in the u.s -- annual death. so, of like two cancers, meet such criteria, so first of all one is a pancreatic cancer and
second is small cell lung so small sell lung cancer has less than 7% of 5 year survival rate and there is about a 30,000 deaths of small cell lung cancer in the u.s. each year. first obstacle is that small cell lung cancer can develop decades after smoking cessation.
although you have lower risk of developing lung cancer risk like nonsmoker as mentioned by the previous speaker. the second obstacle is small cell lung cancer tends to metastasize at a very early cause of disease. and this makes a surgery less
effective in these cancer types. and the third obstacle is although the small cell lung cancer is radiation therapy and chemotherapy, over like 95% of a patient develop a resistance very quickly. and the last obstacle is there is a lack of tumor tissue for
the clinical molecular and cell biology studies this is partially because of the surgery is not a very effective treatment modality in these cancer type. so i'm going to talk about basic formation of cancer biology of the small cell lung cancer.
because of the time limit so it's not impossible for me to talk about every aspect of the small cell lung cancer so i'm going to focus on the genetic abnormalities of the small cell. so the first genetic abnormality of the small cell lung cancer being noticed is the deletion of
chromosomal region 3p21. so in the 1980s, several groups made findings the chromosome 3p is commonly deleted in the small so, on the left side show the standing of chromosome 3 in five different small cell lung cancer cell lines. so on the left side is the
chromosome 3 at the normal length and on the right side it showed the shorter chromosome 3 in the cell lines. and here affectionately it is kind of specify what region of the chromosome 3p is deleted. so, a common region that is shared in the cell lines to be
deleted is chromosome 3p21. so the later study found that chromosomal regions 3p21 is not only deleted in the small cell lung cancer but also deleted in all major type of lung cancer as well as other epithelial so in the precancerous legion, you can also find deletion of
this 3p. i have just lift end to interesting talk given by dr. thomas reed of the genetics branch at nci. so, he has found chromosomal 3p deletion is also present in the precancerous legion of the cervical cancer.
so he is doing a clinical trial to test whether this can be used as a biomarker to detect the precancerous lesions of the cervical lesion at a high risk of serve calling cancer. one point is the the 3p deletion is common in the small cell lung cancer t is kind of suggesting
there may be a tumor suppressor gene in the chromosome 3p that is important for the small cell lung cancer carcinogenesis. and so there is a number of candidates such as backa one and the exact identity of these tumor suppressor genes are still unknown.
so 13q and 17p deletion are also very common in the small cell clung cancer. so, the retinal plastoma gene is actually located on the chromosome 313q. so, the dr. fredrick k asked this question back in 1980s whether rb is lost in the small
cell lung cancer or not. so, here is northern blot taken from a piece of paper. so, in the top part of the picture showed the northern blot results of the rb probe and the bottom part showed the results found beta actin probe. so here the density of the band
is shown abundance of the mra. so the first column is the normal lung cancer or normal lung tissues and the number 2-6 are from the non-small cell lung 7-9 are from the car sinnoids and neuroendocrine tumors of the lower malignant potentials. and from 10-16 are from the
small cell lung cancer cell lines. so, what he observed is there is a loss of rb expression in the car sinnoids as well as in the so he extended these studies to many other small cell lung cancer cell lines and the results it was shown in the
panel b so here in the panel b1-2 belongs to the non-small sell lung cancer lines and 3-12 are small cell lung cancer cell lineups. so you can see here that the rb is lost in most of the small cell lung cancer cell lines. also look at p53, located in
the chromosomal 17p which is also commonly like deleted or in so, what he found is that the p53 can have many different types of chromosomal abnormalities in all types of lung cancers. and here is a table that showed the p53 can be deleted
homozygously or it can have different size of mrna due to the truncation or it can have small point of mutations. so i would like to point out that back then we don't have genome sequencing. so he used the rna's protection assay to detect a small point
mutations in the p53 genes. so, here on the table he is having to specify that h187, h345 and also h378 do not have any detectable mutation. but these days we know that there is also -- there is like still small point mutations in p73 genes.
p53 genes in these small cell lung cancer cell lines. all right. so, as i mentioned, we know that rb and p53 genes are commonly inactivated in the small cell lung cancer cells. but the question is whether these loss of tumor suppress are
genes or these are mutations are causative to the small cell lung cancer or it's just a passenger mutations. so, this is the evidence provided by the dr. burns out of the netherlands that if you delete the rb and the p53 in the airway of mouse cells, it
actually can generate the small so what they do so they actually, injected the adinovirus into the trachea of genetically modified mouse model so they can conditionally knockout the p53 and rb genes in this mouse. so, about two months after the
intertracheal injection of adinovirus, you can see the hypoplastic lesions present here. and this is hae staining and showed the brdu staining showing these are highly proliferated. so if you wait for 6 months after intertracheal or
administration of the adoneo virus, small cell lung cancer then develop. importantly, the small cell lung cancer develop in this mouse model behave similarly as small cell lung cancer in humans. it ma test size very early and easily.
there is a long latent period to develop the small cell lung cancer in these model. the group has tried to combine the p53 and rb knockout with overexpression of myc or other type of oncogenes and that has a greatly accelerate the formation of the small cell lung cancer in
these transgenic mouse model. so, these piece of how to establish the causative row of p53 and rb mutations in the the small cell lung cancer can also develop with carcinoma after treatment with the tyrosine kinase inhibitor. so, this is a case of -- a case
reported by dr. jeffrey engelman out of dana-farber. so, the a depicted over this course over one patient with a long adenocarcinoma. so this patient was diagnosed with adenocarcinoma with egfr sensitizing mutation a50ar in 2008.
so, the patient was treated with erlotinib. about a year later, he developed so the genetic profiling show that the small cell lung cancer has the egfr858r mutation in addition to the small cell lung cancer also has pip 3ca mutation.
so this patient was treated with chemotherapy and radiation. and after the patient finished the chemo radiation therapy, the patient was back on the erlotinib treatment for the egfr positive lung ard no carcinoma, in the middle of 2010, the patient had reoccurrence and
then was treated again with chemotherapy and radiation and also erlotinib. however, the disease was resistant to the treatment and showed that it has adenocarcinoma and the small cell lung cancer at the different sites.
so the patient eventually died and had autopsy and here showed the genomic profile of the different cancers in these so just want to point out here is to look at p53 status. so in the normal liver of these patients, p53 was wildtype so in the adenocarcinoma, there is
deletion in one allele of the p53. however, when patient develops small cell lung cancer, there is a loss of heterozygosity of p53 in these patients. so if you look at egfr mutation, all of the tumors either the adenocarcinoma or the small cell
lung cancer, all have egfr and a50ar mutation meaning that the cell lung cancer all derived from the same clone initially. however, when the adenocarcinoma become resistant, it gained egfrt790 i mutation. but such mutation was not present in small cell lung
instead it had the pick 3ca mutations which was not impressive in the distant adenocarcinoma meaning that these two tumors diverged at a certain stage and then became separate mutations and that becomes resistant to the treatment.
so, here is demonstrated that p53 is lost in the small cell lung cancer derived from the adenocarcinoma. how about rb? we learned from the slides, rb is also commonly lost in the so here in the table 1 it showed a number of the patients the
genomic profiling data and it is probably too busy for you to really look through right now. but what you like to summarize is in the patient who initially had a adenocarcinoma but developed a small cell lung cancer as a specify here as neuroendocrine tumor and there
is a loss of rb in all cases. and but rb is still impact in the tumors that remains to be adenocarcinoma is applied or after the treatment. so, again, this proves that the p53 and rb are important in the small cell lung cancer carcinogenesis.
so now we are in the genomic era of cancers. so there have been number of large studies to look at the genomic abnormalities of the and actually one of the reason to do such kind of analysis is try to find whether there is any actionable mutations being the
small cell lung cancer that can be treated with either target therapy or other kind of small molecule inhibitors. i only include data from the largest study published in 2015 by nature and this study sequenced small cell lung cancer over 150 patients and the data
is summarized here. so on top of these page show p53 and rb1 are the most commonly muted genes in small which is consistent pratt previous finding so mutation are present in like over 95% and 85% of the patients. these two are histotransferase
so those genes are involved in the gene transcription and a certain class of genes commonly mutated in small cell lung canner are notch family genes notch 1-4. on the right-hand side show wagenes are commonly deleted or amplified in the small cell lung
the blue are commonly deleted in small cell lung cancer such as p33 and rb, and fh19 is located in the chromosomal 3p deleted in it shows the next family genes are commonly amplified in small cell lung cancer and the fibroblast growth factor receptor one and the rs2 place a
role in the pi3 kinase signaling are commonly amplified. so here the pathways currently effective in the small cell lung cancer so let's look at the top left hand which i showed that p53 and rb1 both of them are important to checkpoint of the cell cycle are commonly mutated
in the small cell lung cancer. it was found that the cyclin d1 is amplified due to the chromosomal rearrangement so it can by pass the rp1. and the oncogenes are like a very rarely mutatedded in the small cell lung cancers such as 3ca and the oncogenes.
here in the lower right hand it showed the notch family genes are commonly mutated in the notch signaling is a very important in controlling the neuroendocrine variation by affecting expression of asdl1 as shown here. so sdl1 is a monster
transcriptional factor for the newer end crin differentiation. and -- neuroendocrine -- and notch signaling can suppress expression of sdr1. so in the small cell lung cancer when notch receptor is mutated, it decreases notch signaling and leads to small cell lung cancer.
and this is cartoon showing the cascade of the notch pathway. so, the notch receptors are membrane receptors. so when there is a property ligand and the notch receptor will change the configuration so that it exposes the sites to the seretase, while cleaved, the
notch receptor so that the intracellular domain of a notch receptor can serve as the signaling molecule and migrate into the nucleus. so once the intracellular domain get into the nucleus, it can bind to compacts by the protein and as well as partner rbpj.
and i would like to think that notch intracellular domain serve as a key so insert into the lock by the one and activate downstream transcription such as-- gene expression. notch signaling in cell lung the author of nature paper established a transgenic mice that can overexpress the notch 2
intracellular domain. so the notch signaling can be turned on in these transgenic mice. so here show the formation of small cell lung cancer in the transgenic mouse model with three genes knockout including p53, rb1 and rb like 1.
and here shows the transgenic mice with the forced expression of notch two intracellular domain. you can see from here that there is a decrease of number of small cell lung cancer in the lungs of these transgenic mice when the notch 2i cds overexpressed and
here shows the quantitative data that the number of the tumors as well as the volume of the tumors are decreased and on the right-hand side showed the survival curve of these transgenic mice and the forced expression of notch two intracellular domain extends the
survival of the transgenic mice. so, again, it's support data and the enactivation of notch signaling plays a very important role in the carcinogenesis of so, next i'm just going to give you a few examples of the successful translational medicine in the small cell lung
so first i want to talk about rova t. so, again it's related to the notch signaling. so, let's kind of go back and look at these segment cascades. so there is a number of the ligands that can bind to the notch receptor and the role of
the ligand is dll3 which is a ligand that can inactivate the so, a group of scientists from the -- a company based in the southern california made a finding, made independent finding a number of years ago that the dll3, which is antagonizing ligand of the notch
signaling is overexpressed in here shows the expression of dl3 based on rn seq data. so the y axis showed the kn number and so as you can see from here, the small cell lung cancer has a much higher expression of dl3. about 40-50 fold higher than the
normal lung tissue and here shows the small cell lung cancer cell line which also has pretty high expression of the dll3. so, i just want to stop here for a little bit. so if you have a such kind of finding, how would you like translate into patient care?
so, in the scientists found the development of a new strategies to target the dll3, the small cell lung cancer cells. so they develop antibody conjugate and shows structure of the antibody and this is something they call rova t. this can recognize dll3 on the
surface of the small cell lung cancer and this antibody is actually conjugated through a linker with the very toxic chemo agent called pvd. it can bind to the smaller groove of the dna damage but pvd itself is a very toxic so you can not give it to the patient.
so, what happened is that wops you put the pvd as a payload to the antibody and then it becomes a very target agent that can deliver the pvd specifically to the capser cells that express the dll3. so, here shows the experiment data of the advocacy of rova t
in patient derived xenograft mouse models. let me just help you to go through these. so ep chemotherapy. so there is a tumor volume but what happens is that the tumor return very fast. and then if you don't give any
treatment, depicted as tumor will just grow continuously and here the mice give dll3 antibody drug coverage gate and tumor size shrink again but importantly, this affect lasted for a very long time most better than chemotherapy amount. so here they also kind of tested
the rechallenging of the chemo again, those mice were treated with a chemotherapy first and then when they had reoccurrence, these mice were treated with the ep chemo agent again. as shown here, this time the response lasted each shorter comparing to the first time and
those tumors are growing back as depicted with the blue curve so dpl3 antibody drug conjugate seems to be more effective than rechallenging the pdx mice with chemo agent again where the tumor had reoccurrence. so, with that data then they moved this drug into the
clinical trial and this is the result from the phase i trial recorded by memorial sloan-kettering in 2016 asco meeting. so in this phase 1 trial. they included 74 small cell lung cancer patient who had a reoccurrence after the one line
of chemotherapy. so some patient has sensitive disease and some had resistant and refractary disease. definition of sensitive or resistant and refractary disease are explained here. so if a patient does not have a progression disease after three
months after the first line chemotherapy, then it is considered sensitive disease. so meaning that disease is sincative to the chemotherapy and in fact is durable more than threes month. if a patient had a reoccurrence or progression of disease within
the three months after considered resistant disease. so if a patient did not have any response but continued to have a progression disease meaning that chemo continues to grow even when patient isn't receiving chemotherapy, it's considered as refractary disease.
so we just wanted to pointed out here that about 67% of patients have like measure 50% of tumor cells with dl3 expression membershipping highly expressed over the phase i clinical trial. so as shown here, objective of response rate that includes completed response meaning no
more tumor plus partial response that means tumor decrease size by more than 30%. so if you don't select biomarker in all 9 patients no matter whether patient express dl3 or not, you have 18% of patient who had response. when you select for patient who
has more than 50% tumor cells expressing the dll3 that is response rate improved to 39%, which include patient who had response, and patient with the stable disease after the treatment, it showed that when you select for the biomarker response rate got much better.
this saturday water plot of the response at the treatment and here the patient is based on the expression of dll3. so all these patients had a shrinkage of tumor but the patient under the line is considered have a response after treatment.
most patient with response had expression of dll3 of more than 50%. and some do not have the data on the dll3 expression. in order for the drug to take effect. the company has moved forward to gain approval by the fda for the
drug. so, now we are also in therapy so it is impossible not to mention about therapy in the so this is famous data that shows mutation load in the different cancer types and because of the time limits, just want to point out that so as it
is shown in by the arrow here, the small cell lung cancer also has a very high mutation load compared to other tumor types. so, what is the significance? so although we don't really know exactly what is relationship between the mutation load and the response to the immune
therapy but it appears that if a patient with tumor has more mutation, so that generates more new antigens that can be recognized by the system in the body. so that patient may have key lymphocytes that can recognize the tumor cells.
so although the body can have the t-cell that can recognize the tumor it doesn't mean the immune system is turned on to eliminate the tumors. that is because the body has the different immune checkpoints that help to inhibit the immune response.
that is used as a protective mechanism to protect the body parts from the -- being attacked by the immune system. the tumor cells can take advantage of that. so, there is a two immune checkpoints that can be targeted by the approved fda drug.
one is the ct i-4 that can be targeted and the other is pd-1 and pdl1 system so that can be targeted by the pdl1 or tb1 antibody. so as shown here, the pd-1 or pdl1 positive cells are in the stromal cells instead of inside tumors.
but when this paper was published in 2015, there is a letter to the editor submitted later that shows if you use a different antibody, some of the small cell lung cancer can have positive staining of pd-1 or pdl1. so this is the summary of that
and basically, it did not find any pd1 or pdl1 positive cells in the tumor but found a positive cell in the stroma. so, it raised a question if you give a patient just pdl1 whether that will be effective enough to activate immune system because it seems like immune cells are
like out of the periphery of the tumor but not really inside the tumor. so there is kind of a these two, this idea or two combined the anti-pdl1 with ctl84 antibody. so this is a checkmate 032 study and so here they tried to like 3 different arms.
the first arm is just anti-pd-1 itself. it's elube mab and the second and third arms are combination of anti-pdl1 with ctla4 antibody as different dosing frequency. so the middle one is the 1 plus the 3 and the last one is the 3 plus the 1.
so this is patient characteristics. so just want to pointed out if you look at pdl1 expression level, the most has little expression of pdl1 and here is the response data. so if in the arm there is 10% response rate and the if you
combine this with ep lube mab so response rate doubled and on the right-hand side which is the spider plot which shows how each patient responded to therapy. so there is a long tail which show that the tumors can be put in check by the therapy for very long period of time.
and the therapy itself is not benign. so very severe autoimmune disease after treatment. so there is a number of other promising agents that are under clinical development such as b1 inhibitor and 20% of objective response rate was recorded in a
small phase i trial. powerful inhibitor is promising and so a 10% the popular response was reported in phase i trial. and identified as biomarker of the 1 inhibitor in small cell and kinase sample promising in 21% of partial response rate was
reported in multi-center phase ii trial. a large study screening over 64 lines and over 100fda approved drugs as well as over 400 investigational drugs and look to see what drugs is effective in the small cell lung cancer and this is a paper that you can
look. and so the typical response is that some tumors are sensitive and some are resistant and this is to be the common theme in all these drugs. and this is a website that you can look at their data because it has the gene expression data
over 64 cell lines and also includes the ic50 values of over 500 compounds so it can be a great resource if you're interested in small cell lung cancer research. because of time i will skip these. so just to summarize small cell
lung cancer is cancer and new therapy is needed and p53 and rb1 are commonly lost in the small cell lung cancer and the newer therapies such as antibody drug conjugate and immune therapy are common. thank you very much. and if you have any questions, i
will be happy to take them. >> subtype. i think so. yes. so, in the study, she tried to group small cell lung cancer based on the microrna expression and that seems to be -- in expression of
microrna may correlate with the sensitivity to a particular type of a drug. so that seems to be the case.
Tidak ada komentar
Posting Komentar