Cervic Cancer

>> good afternoon. we have some very exciting lectures today. the first is john schiller, he got his ph.d. from the university of washington at seattle, and then he sort of came to nci, worked his way up through the ranks, became a

postdoctoral fellow in 1983, senior staff fellow in 1986ing senior investigator in 1992, section vaccine, vaccining to prevent oncological hpv infection. john? >> thank you for attending. it's hard to give a lecture

if there's not somebody here. now we're cool. thank you for being physically present in the audience, not just in videoland because it's hard to give a lecture when there's nobody sitting in front of you.

if you have questions, feel free to interrupt, those of you in videoland if you have questions my telephone number is -- no, i'm not going to do that. i'm going to tell you today about a relatively brief overview of the association of hpv and cancer,

epidemiologyically and biologically and spend time on something i've been studying, development of prophylactic vaccines, why they work, the history of that, and then the implementation issues where this is an issue where now we have something that's a good

intervention to prevent a series of important cancers and how do you get people to use it? i think many of you may realize preventive measures are better than therapeutic measures, you heard about someone that has cancer, they are motivated, but with prevention people don't

know who is going to benefit and who isn't. so infectious agents are the cause of many cancers. it's been estimated approximately 16% of all human cancers could be attributable to infectious agents, the big three are here, hebobacter pylri,

human papilloma, and hepatitis b and c cause half a million cancers every year. the important thing, the important aspect of cancers associated with infectious diseases i'll be bringing up at the end, most of the infectious associated cancers occur in the

less developed world. this is true of hpv where you can see the number of cases in the less developed world, in comparison to more developed countries. we think about hpv associated cancers worldwide we focus on cervical cancer because it's

both got the highest number of actual cases and also the highest attributable proportion, virtually 100% of cervical cancers are initiated by hpv infections and it causes other cancers as well, 5% of all cancers in the world from hpv. there's different times, over

100 types of hpvs, two types predominant, type 16 and type 18. this is true in parts of the world that you look at. types 3, 4, 5 and 6 vary and are the same clusters but the rank order changes as you can see here.

if you look at the united states, cervical cancer doesn't dominate the spectrum quite as much as it does worldwide. there's two reasons for this primarily. first of all, pap screening reduced the incidence by greater than 80% so we're down.

there's been an epidemic of hpv positive oral pharyngeal cancer in two decades, increased two-fold. it makes up a substantial contribution to the hpv-associated cancers in the united states, it's predicted within 20 years to be more hpv

associated than cervical cancers. the other thing that this causes is that it means that male associated cancers make up a larger minority than they do worldwide because oral pharyngeal cancers are more common in males than females.

the types of hvps that cause cervical and other types of cancer cluster over here, we talked about another group that caused genital warts, which are distinct when the types that cause common hand and foot warts or a whole other group that mainly cause asymptomatic

infections in all of us. hpvs have a peculiar lifestyle related to vaccine development, and they are the only virus, they are a virus that only replicates in squamos epithelium. if it would transfection any other tissue it wouldn't produce

virus, in fact it doesn't. the reason probably for doing this is that it allows it to evade immune evasion. we think the virus has to get to the base of the membrane to start infection and affects basal epithelial cells with low level of expression of the early

proteins involved in autonomous replication transcription and ultimately cellular transformation. and then only as the cells become differentiated do we get large amounts of proteins expressed, particularly the virion proteins and release of

infectious virions. this area of the epithelium is not normally subject to strongly immune surveillance, these are basically dying cells. so this allows the virus to persist for quite a while before the immune system eventually recognizes and eliminates it.

so basically productive hpv infection involves hiding in plain sight. this is a high risk hpv infection on the cervix. it's very inpatient, it doesn't even cause hyper proliferation. again, this is an example of a marker of early gene expression,

one of the oncogenese e-7, it's the e-7 protein made of low levels, you get amplification of the signal and expression of mcf, and then the late genes, associated with the virulent, are expressed in the upper levels of the epithelium. we have a very good idea how hpv

is associated with cancer on a growth level. the mild lesions that are diagnosed in a pap smear as low grade squamos epithelials, or in histology as carcinoma, great 1, as the virus causes differentiation there's no more virus production, so it's an end

for the virus and host. you get higher levels of e-6 and e-7 accompanied by viral integration leading to a series of dysplastic changes from intermediatate to high grade dysplasia and the entire epithelium is a proliferative compartment, it's carcinoma in

situ leading to invasive cancer. why does the virus encode oncogenes? i told you it makes no sense because the progressive lesions don't produce virus. it does it as a by-product of overcoming a problem because as i told you, the virus replicates

its dna in terminally differentiated cells that are not themselves dividing. so it's got to figure out a way of tricking the cell into thinking it wants to divide to make the replication machinery that's needed to make its own genome.

so it's got one protein, e-7, that tricks the cell into aberrent proliferation. if you get abnormal signals for replication, this is the signal for the cell to undergo apoptosis. another protein prevents apoptosis so the combination of

inducing aberran proliferation leans to genomic instait about. e-6 and e-7 interact with a wide variety of proteins. all sorts of cellular partners, neither one has enzymatic activity, and a bunch arist willed here. still also unclear which of

these activities are important for carcinogenic progression and which are important for normal life cycle. what's clear, one of the interacting partners of e-6 is important for carcinogenic progression, and that's interaction with p53.

interaction with p53 binding with e6ap which leads to degradation. the same is true of e-7. one of the important interactions is likely with r.v. and p-107 and p-130 because changes in p53 and r.v. by mutation are very common in

cancers that aren't associated with hpv but again this interaction as well as interactions with cyclins are important in driving the cells. another feature almost unvariably it occurs as a particular site. the female reproduce tract at

the cervical transformation zone goes from being stratified epithelial to a single layer. we're not sure why they arise a recent paper, it was shown in a very small portion right at this transformation zone there's a very small layer of cells that maintain embryology cal

phenotype including keratin 7 normally expressed in fetal cervix but not expressed on mature cervix so we think there's something about these cells that make them especially prone to cervical carcinogenesis, if you get an infection here or here, it's

much less likely to undergo carcinogenic progression. this transformation zone is important because some tissues have this and other ones don't that hpv infected. there's a cervical transformation zone at the cervix, there's also a

transformation zone in anal epithelium where most cancers arise. hpv infections of the vulva, vagina and penis are common but cancer is rare. why is it these areas are the place where cancers arise? which is to note an example of a

high grade lesion, cin3, and the fact that generally a high grade cursor takes at least ten years. schematically this is the time line to cervical cancer. infections are extremely common but most of them go away spontaneous lie. the lifetime incidence of

genital hpv infection is 85% in the united states, being sexually active is synonymous with having hpv infection but most go away spontaneously which eliminates the risk of cervical cancer. it's not clear if they go away entirely or become latent.

the immune system keeps it under control and now if you can't detect the virus, the chances of getting cervical cancer from that virus is extremely small. so what's on the causal pathway to cancer is persistent abduction, the high risk type, specifically 16 and 18, this is

by far the most important risk factor for progression to pre-cancer takes a decade, and then some of them go away, but then these pre-cancerous lesions as i mentioned within a decade oftentimes advance to cancer, so cervical cancer is unusual in that it actually affects

relatively young women. women in their 30s, 40s and 50s, by their mid-30s to 40s the peak is flat and it's not predominantly a disease of older people like many cancers. so they're not more common when you start to get into the 60s, 70s and 80s.

again, it's probably due to the viral etiological -- etiology. i'm going to turn to what we did to develop a vaccine against this virus. i want to take you back to the time when we first started to develop this vaccine which is the early 1990s.

at this time, what we knew is that if you took -- if you did intramuscular inflection of papillomavirus virulent in animal models, this could induce protection. you could take serum from an animal that was injected with these in transfer, we thought,

hey, it's antibody mediated, but when we tried to mimic this with in vitro-derived protein, denatured form of the major capsid protein, l1, for instance, made in e. coli, or l1 synthetic peptides, e. coli-derived peptides didn't protect at all.

an immunogen was the critically correct feature. when you can't reproduce this in large scalable quantities, there was no source of authentic virulents to make a vaccine as you might for measles. the vaccine would contain these viral oncogenes, you would never

be able to use a vaccine that could potentially contain oncogenes in healthy young people, the vast majority of which were not destined to get cervical cancer. what we did is generated what was called virus-like particles, and we injected -- we introduced

l1 into a baculovirus expression season, we use that in production in insect cells because it had already been approved for human clinical trials, at least other types of proteins generated in this system. we found this one protein alone,

l1, assembled into virus-like particles, vlps. we injected vlps, they acted like real virus, high titers of. they didn't contain e6 and e 7 they were noninfectious and oncogenic, providing a source of vaccine to test in people. when we tried to develop this

there was a lot of skepticism that a vaccine against sexually transmitted disease could work. we talked to a whole bunch of companies, most of them said, your data looks great, all these great neutralizing antibodies but we know sexually transmitted disease vaccines don't work.

two companies took a leap of faith that maybe this could work, glaxosmithkline and merck and independently developed vaccines, which are similar in concept, based upon virus-like particles, but somewhat distinct. cervarix not used much in the

united states protects against 16 and 18 which i've said is the cause of 70% of cervical canner is. they make it in the same production system as we did until baculovirus-infect the insect cells. the merck convenience also

contains 6 and 11 responsible for genital warts, made inest yao. both are three doses, and this shows you the time line of developing this. one thing i'll leave you with, if you want to develop public health interventions, you better

be patient. it takes a long time. this shows in 1982, hpv16 was discovered, in 2008 they got the nobel prize. by the time the convenience got licensed it was 15 years later. this was with two companies, two of the best vaccine

manufacturers in the world, competing with each other to bring this vaccine to market to capture the market, and it still took 15 years. because what we're preventing is initial infection it will take another decade before we see a significant drop in cervical

cancer rates based on the vaccine. these types of interventions especially based upon prophylaxis, prevention of cancers, takes a long time to see a benefit. one of the things that's important to point out is that

it was absolutely critical to understand this process of how you go from virus infection to cancer in order for us to have this vaccine today. because we could have never had this vaccine licensed with an end point of cervical cancer for two reasons.

one, it would have taken decades and decades, you had to follow the women just about forever in order to show that you protect against cancer, and secondly, income a trial with active follow-up we couldn't let anybody go on to cancer. we had to do careful screenings

of these women by pap screenings and by measuring their dna content of their cervix so we would identify pre-malignant lesions and remove them surgically before the women went on to cancer. it would not be ethical to allow them to get cancer, the basic

understanding of how the cancers progressed from other lesions was important because for licensure, prevention of cervical cancer, we only had to show that it protected against development of the moderate and high grade dysplasias by the 16 and 18, the two types that were

targeted in the vaccine. based upon that, the fda says, okay, we know, we believe if you prevent that, it will prevent and there were a series of trials done both by america and gsk and the national cancer institute here that looked at the efficacy of vaccine.

if you look at the most cancer proximal end point, ciniii, both vaccines gave 100% protection against disease caused by the vaccine targeted types. so this was remarkable. no one thought this this could work, let alone work with 100% efficacy against the types that

it's targeting. it was also strong protection of gardacil, protecting against genital warts, 95%, and in males somewhat less, we think the main reason is because the vaccine prevents infection but doesn't treat infections once they occur.

i think the idea is we probably miss more prevalent infection in men, in the male genitalia, than at the cervix. so because of this, there was more men who had infections that we missed, and they emerged during the process of vaccination, and looked like

they were incident infections but were prevalent before we vaccinated them. if you look at the different types of cancers that are associated with hpv and you look at have we demonstrated protection from infection, upal neoplasia, we have not done this

for penile cancer. we didn't see enough cases in the trials to show we protected against lesions although we did protect against infection. in the case of oropharygnea we haven't identified the pre-malignant lesions and have not determined it will protect

against the lesions, how it leads to cancer, before we can do those studies. it doesn't prevention invection or disease caused by other hpv types that cause 30% of cervical cancer and doesn't induce regression of established infection or prevent progression

of hpv-induced lesions. it doesn't act therapeutically. i've mentioned that hpv 16 is targeted by both the vaccines, 70% of cancers. there's a new vaccine on the horizon which merck is developing, it conveyance the five next types that are

associated with cancer. each one of these contributes relatively small amounts, this could presumably go to protecting 70% of the types to 90% of the types, and they were able to show efficacy trials, this efficacy trial they couldn't use a placebo control

because it would have been inethical. they compared gardacil, and showed that the immunogeninicity was the same. the new types that were being targeted, in comparison to gardacisil, it prevented 96% of cin 2/3.

they applied for licensure to the fda and it's likely this vaccine will be able i would say in the next six months, in a year at the outside. merck hasn't been popularizing this much because they don't want to stop sales of the current vaccine and have people

wait for the next vaccine but it's likely within a year this new vaccine will come out. so if you have a 12-year-old daughter or friend who has a daughter, i would hold off for a few months and get the new one of the interesting questions, what will happen with

people who got the old vaccine? will they want to get the new vaccine? will it be recommended? it raises implementation issues. we think it protects from initial infection which again is surprising because a lot of prophylactic vaccines, hepatitis

b vaccine, doesn't necessarily protect initial hepatitis b invection but prevents you from getting the disease so it nips the the infection in the bud but we think that this vaccine has sterilizing immunity because most vaccineees never test positive so if you swab the

cervix you never see positivity by pcr. grade ii appear early in the trials which suggest most are emergent prevalent infection so the first six months or year is where you see most of the breakthrough infections, after that essentially no one is

getting infection out to 8 or 10 years now. so we think that it's emergent, prevalent infection measured early on. antibody response goes up and then drops. we think the antibodies are likely to be the immune

mediator, we generate high titers, because cross protection mirrors antibody mediated cross-neutralization and we can transfer protection in animal models by taking serum and putting it into the naive animals, challenging naive animals with the antibodies in

that animal to be completely protected. the other point is that cell mediated factors function after infection occurred, getting sterilized in immunity. so we think that the neutralizing antibodies are recognizing multiple distinct

epitopes that vary from type to type. probably this type speciesation occurred in response to neutralizing antibodies, occurred over millenia. epitopes are recognized on the vlps so if you get rid of for instance the de loop here you

still have antibodies against the hi loop and so that you would prevent infection that way. we think it's unlikely the virus will rapidly evolve to escape neutralization. this is different from hiv. they evolve as rapidly as dna

because they replicate themselves. the antibody response is stunning. everybody responds well to this the very few people who don't make a good response to the vaccine, and the response starts very high and then drops as your

short-lived plasma cells die off and you're left with long-lived cells that pump out antibodies for as long as we've been able to measure, 8 or 10 years. between year 2 and year 8 or 10 there's no difference in antibody responses. we think this is going to

continue out long term. so we don't know the duration of a protection, clearly now we're vaccinating 11 and 12 year old girls, more and more boys. we need the vaccine to last for a long time. some people criticize, how do you know it's going to last long

term? so far the antibody titers have not changed between year 2 and 10, there's no reason to think they will fall off the table at this point. we're optimistic this is going to induce long-term protection through the years of greatest

risk when people tend to change sexual partners the most. just to show you how great this vaccine is, inducing antibody responses, the costa rican trials, some women got one dose. they were not randomized but some got one goes. a vaccine never worked after one

dose. all subunit vaccines are given in multiple doses. after one dose we found by six months, the antibody titers stabilized. we have unpublished data, six years, there's no change. no women changed by more than

two-fold, about the limit of detection of our assay. antibody titers stable buysed but all women remained serum positive at levels nine-fold over the levels from natural infection, you get very little bang for your buck for three doses.

none of the 150 women who got one dose exhibited hpv 16 infection versus 15 in the controlled. this is quite robust data, look at the confidence intervals. we think this could be the first vaccine to work after a sippingel dose which would be

revolutionary, in terms of implementation could make a huge difference. we want to do a formal randomized trial to compare one, two, and three doses and see what happens. we think the way the antibodies get there is by two mechanisms.

so as i said we're doing intramuscular infection, maybe those who know immunology, systemic immunization is a lousy way of -- at the cervix half the antibody is igg, it's the same in the lower respiratory tract, not the upper. it's thought this involves

transudation. and we can measure this antibody floating around in the mucous but it's at levels about 10 to 100-fold lower than what's in the serum. in order to start infection, the virus has to bind with the membrane before it can be

transferred to the cells, allowing for exudation that is interstitial into the wound. near the site of infection it may be close to what's in the serum which is quite high. to conclude, the convenience are highly effective,gardasil is

also highly protective against genital warts in women and men. protect is type-restricted. and the duration of protection is unknown, but strong protection of 8, 9, even 10 years after antibody levels have stabled, we think these vaccines have great potential to prevent

the half a million or more or 600,000 cancers every year that are associated with hpv but they only have this potential if they are used. this brings up -- oh, okay. i'm getting ahead of myself. hpv vaccines are now established products, they hold more than

100 million doses delivered, 25 million of cervarix. cervarix was about a year later and gardasil in 2009 for genital warts, anal cancer in 2010. they are licensed around the world but that was a difference between licensing and incorporated into the national

vaccine program and being made available to the majority of the people in the country. most countries, especially emerging countries and developing countries in lower development countries, the vaccine is very grossly underutilized.

happening fairly recently is that some countries are going to less than three-dose programs particularly two-dose vaccines given at zero and six months approved in the e.u. if you give two doses at zero and six months in younger kids, so 9 to 13 year olds, you get a

noninferior antibody response, giving three doses in 15 to 26 year olds, the target of the vaccine trials, this is something that's been shown for the vaccines by the time you go to adolescence, reach puberty, the immune response started to decline, this has led to

two-dose schedules being implemented, not in the united states yet incidentally, although we think we should be probably going to a two-dose vaccine implementation. this leads to the implementation issues. the first question is who to

vaccinate? overall in most of the world girls suffer the majority of the hpv-associated cancers we want to focus on vaccination of girls. after that, there is reason to vaccinate boys, boat because they are the vectors that

transmit it to girls and they are associated with a certain subset of hpv-associated you want to vaccinate young women if you have the money because although by the time a woman reaches her mid-20s there's a good chance show was already exposed, there's a

subset that won't have been exposed so it could still benefit. older men would be the last group because they are least likely to benefit. now, it's important we vaccinate before the onset of sexual activity, and the reason for

that is pretty dramatically shown here. this shows the time since first intercourse, and the cumulative risk of getting hpv infections. you can see by one year, by four months, 20% acquired infection. by 12 years it's 60%. you need to vaccinate before

they become sexually active because once they have the infection, the vaccine won't do anything to make it go away. fairly recently the cdc has gone with the recommendation that boys 11 to 12 should be routinely vaccinated, and before this the vaccine recommendation

was permissive, which means you could do it if you want to do it but it was basically telling insurance companies they didn't have to pay for it. this change was made because as i'll show in a minute, uptake was low in girls and this was about the time when we see the

indication of the prevention of ana cancer in boys and girls, it was hard to say we should not be vaccinating for boys, if they could prevent a specific cancer associated with men. what are the issues, how do you deliver three intramuscular doses early adolescence?

there are very few intervention programs targeted at adolescents, certainly not three visits. the uptake rate of adolescent young girls in various countries, you can see quite clearly schools, countries with school-based vaccinations,

uptake rate is good for one and three doses. in the united states and france we don't do a good job. in the united states, it's about 54% get one dose, only a third of girls get all three doses. we're starting to look at this 50 percent one dose a little bit

different than some people. a lot of administrators are looking at one dose as a failure. given what we're seeing, i'm starting to look at it as a positive thing. and overall, the incidence of hpv 16 and 18 infection in the

united states dropped 50% in this age group that's targeted for vaccination in the next few years. and so people are thinking well, there's some immunity because only 33% got full vaccination but it means that maybe even one dose can give you protection.

if you look at the uptake of the hpv vaccine shown in blue for females, red for males, it started to uptick for boys when it became fully recommended by the cdc. if you look at the uptake in women over a longer period of time, you can see it lagged

behind that of two other vaccines, diphtheria, tetanus, pertussis, and meningococcal, not recommended as strongly by pediatricians and others as other vaccines. doctors say it's time for you to get your dpap vaccine, and for the hpv vaccine they say this is

available if you want it. and just that difference in messaging that's they are giving is going to account for the there's a disincentive on certain parents' part because they think it promotes sexual activity and things like that which is shown not to actually

be the case. that's too bad that we're not using this vaccine more in the united states because like in australia where they have 80% coverage rate there's been a dramatic decrease in the population in intermediatate and high grade cervical dysplasia in

the vaccinated age groups,oeven a little bit in this age group but not in the older women not being vaccinated. we're seeing populations effect from these vaccination programs. because genital warts is an earlier manifestation of hpv infections, that is high grade

dysplasia, you see a more dramatic effect so the decrease in young girls, the youngest age group is decreased by 92% since vaccination started in 2007, really amazing, and even in the next age group by 73%, perhaps even more remarkable is that among heterosexual men which

weren't being vaccinated until 2011, you can see that in the youngest age groups obviously having sex with the girls that are being protected you can see that there's been a drop of 82% in genital warts among young men who weren't being vaccinated at all.

this is a clear example of a very dramatic herd immunity effect. we have to consider the effect of vaccine on cervical cancer screening recommendations. cervical cancer screening reduced cervical cancer rates by 80%.

the last thing we want to do is use the vaccine that protects in 70% in abandoned screening doing better than the vaccine. we have to convince people, women, to continue with their screenings because the vaccine that won't help with established lesions and it's type restricted

and won't be expected to prevent 20 to 30% of cervical cancers but the new vaccine is up to 90% so it's better. if you look at the cost of screening, it accounts for most of the direct cost of intervention to prevent cervical cancer in the united states.

cervical screening accounts for 82% of the cost of prevention, much more than the treatment of cancer, so we want to go to a screening program that makes more sense for vaccinated women in the long term. and basically what we hope to do is shift to an hpv based

strategy, vaccinating young girls to drop this peak of high risk hpv infection and replace the pap smear which is relatively insensitive. it doesn't have what's called a very good negative predictive value. if you're negative for pap

smear, there's a reasonable chance that it missed your high grade lesion and your chances of going on and getting cervical cancer in ten years is substantial. hpv has much better negative value if you do a swab and you don't have hpv dna, the chances

of you getting cervical cancer in ten years is really, really low. so what we hope to do is replace this more sensitive test done for this test which is done very often. the combination of vaccination and dna testing, it's been

estimated could be delivered as much less -- much more cost effective than doing pap screenings multiple times throughout a woman's life. lastly, delivery to economically disadvantaged women, 80% occur in developing countries, dealing with this issue has got to be

multipronged. both companies committed to sales to gavi at less than $5 a dose, an institution that buys vaccines and delivers it to 72 poorest countries in the world. even at $5 a dose this is more than all of the epi vaccines combined in terms of price.

one of the solutions vaccine manufacturers in emerging countries are working with manufacturers in for instance india and china and brazil to try to produce the vaccine at a lower cost. this is essentially what's occurred with hepatitis b

vaccine when it was first introduced was somewhere up around 75 to $100 a dose and now it's made in india for unicef for 18 cents a dose. this is what we want to do but shorten the time line in which it gets produced in emergent delivery of a few doses will

save money and we're on the path to delivering two doses, and ultimately based upon the data the future of this vaccine is one dose, and lastly we're trying to develop second generation vaccines that could be manufactured in delivery at one of the ideas is make a

recombinant measles vaccine, so you get measles protection and hpv for nothing. on the hoes other than is the merck vaccine, should be recently available. working on vaccines based upon l2 which might give better cross protection and look for

manufacture in emerging countries and the other area i'm not going to talk about but there's interest in is develop therapeutic vaccines to treat persistent infection in pre-malignant and intraepithelial neoplasia.

what about others? there's no vaccine for h elobacter pylori but this can be treated by antibiotics. this affects the lining of the stomach. there's no vaccine on the horizon. i told you about hpv.

hepatitis b, hopefully a lot of you already had that. there's no vaccine for hepatitis c, rapidly evolving by hiv, and there really isn't i would say a promising candidate on the horizon although people tend to work on it. the big thing is that

virologic-specific antivirals will revolutionize treatment of hpv, they are expensive but hopefully prices will lower and we'll be able to treat hcv persistent infectious. holiday and kaposi's sarcoma, most cases are caused by hiv infection, but

when you use harp, people are now on harp therapy, that's pretty much eliminating koposi's sarcoma, and the other cancers there's not enough cases, mostly in developing worlds, there's no interest in developing vaccines for them. i'd like to acknowledge my

collaborators doing some of the work i've described, especially doug lowy and collaborators, especially who were in charge of the nci-sponsored clinical trials in costa rica. with that i'll stop and take any questions. [applause]

yeah? [off mic] they are not more susceptible to infection but carcinogenic progression. we've looked to see whether they are more susceptible to infection and they are really not.

like i say, infection can occur through the entire female reproductive tract but if you do get infection they are more susceptible to progression and we don't understand why. we think broadly the fact that it's more embryological, more stem celllike, that may be why

but that's hand waving, it's not mechanistic. [off mic] >> yeah, there's controversy. the durability is the same but you read some in the literature that it's not as good for 18, for gardacil than the merck vaccine, part of the reason is

it's technical but the assay they use only measures a subset of the antibodies that could be protective, and they have to measure a subset that's not a dominant subset in a lot of women. that goes away but they have other ones that are stable.

there's a little controversy in the literature. gsk wants to make a big deal and say, yes, the protection against 18 is dropping but if you look for protection there's no sign that gardacil is not protective long term against 18, but that's a good question.

two answers. one answer is -- the question is will we get type replacement, one the other types will replace it. the other types are not as oncogenic. given infection, even persistent infection, chance of going on to

cancer are much less with the other type. you would be replacing it by a less oncogenic type. multiple infections are common, but chances of one time being acquired or going away doesn't seem like it's influenced whether you have another type or

two other types or three other types. they seem to be different players at the cervix. now with the impending vaccine, now you're talking about the first seven types, now you have nothing but a couple percent, so that problem really is never

going to arise. >> our next speaker is david salomon, ph.d. from state university of new york at albany, was a fellow at the roche institute of molecular biology, came to nih as a staff fellow and currently he's at nci, in charge of the tumor

growth factor section, his title is stem cells. >> thank you very much, terry. what i would like to do today is give you sort of the conceptual overview of what cancer stem cells are and really it's defining a functional entity, and these are very dynamic cells

that are in flux, they are very plastic, and they interact with their environment in very unique ways, it's important for determining their function and their mobility. and i think before talking about cancer cells we have to understand what are normal stem

cells since cancer cells obviously arise, we think, from either stem cells or their immediate project any, progenitor cells. there's a good deal of interest in the lay population in stem cells because of their potential use in treating diseases,

regenerative medicine, obviously in cancer itself. now, there are must be sources of normal stem cells in the body. the the bone marrow being the primary source in our body, and we saw stem cells have been used to treat various blood diseases,

cancer and m.s., but there are other other sources for example in cord blood, stem cells exist, and they have been used again in a setting to treat various diseases such as stroke and multiple sclerosis or have the potential for treating those diseases.

there are ips cells which are generated from adult tissues, differentiated by genetic manipulation and they have a great potential for use in regenerative medicine. obviously neural stem cells in the treatment of spinal cord injury and some diseases of the

brain like parkinson's disease and alzheimer's and diseases of this sort. so i think when we talk about cancer per se we have to understand cancer is the flip side of normal development. it's abnormal development so it's use of embryonic signaling

pathways in an abnormal context, the wrong time and the wrong place. it's a dr. jekyll and mr. hyde. we have to understand normal development and pathways that control it and stem cells within the process of development before we can hope to ever

approach to treat cancer stem cells or at least know what they are doing. and seminal observations were made many years ago by people like beatrice mintz and barry pierce, onco-geny partially recoo

pip late -- recapitulate. if you took tumor cells from somatic tissue from adult cancer and inject into appropriate environment like the mouse blastocysts, they could become normal. the environment could reprogram the tumor cells.

that was an initial observation, you know, 30 or 40 years ago. recently, the importance of the microenvironment or the niche was made by beatrice mintz and gil smiths, you could expose them to normal embryonic cell environment or place tumor cells in adult tissue stem cell

niches, so again supporting the notion that the environment can redirect a cancer cell in a potential stem cell. finally, the flip side can occur, if you take let's say tumor -ells in vitro and take the condition media from the cells and you treat adult tissue

stem cells that are normal or genetically manipulated stem cells, they will acquire the properties of cancer stem cells or tumor-initiated cells. there's a strong dialogue going on between stem cells and their environment, which will come out in a few minutes.

now, normally stem cells in development exist during early embryogenesis in the epilast layer, the primitive cell that gives rise to germ layers in the embryo, all these give rise to organ systems in the body. we might be able to understand some of the aspects of signaling

in cancer cells, cancer stem cells, which is a very early stem cells. one of the processes which is important in this bifurcation of a potent cell such as epiblast is epithey'llal mesenchymal transition. mesenchymal to epithelial, the

reverse process, occurs in development of the organs from which the germ cell lines give rise. embryonic are the most pluripotent. they have been established from the inner cell mass or epiblast

which is generated from the inner cell mass and these cells in vitro will differentiate into all the tissues of the body under conditions and people are working out signaling pathways that are important in potentially channeled these cells down to various germ cells

that can give rise to some of the different types of adult differentiated cells and this has a great potential for the treatment of diseases and specific pathologies, just listed here, there's a slew of them. they share a general itic

program or transcriptome, listed here is an analysis for 47 studies of human adult embryonic stem cells that have defined common genes expressed in these early undifferentiated pluripotent stem cells. i call your attention to three of these genes, otherwise known

as the trinity. these three genes are the master pluripotent potential control genes involved in keeping keeps in an undifferentiated state pluripotent and can self renew. now the genes are highly complex with respect to how their expression is regulated.

they cross regulate in a complicated way so there's co-cross-regulation of each of the genes, each one by the other two genes. they themselves orchestrate or regulate in a positive manner genes shown here and suppress a number of genes that are

involved in promoting differentiation. so these three genes are important in embryonic stem cells, i'll show you how they are important in cancer stem cells. these genes along with myc and klf 4, i place klf 4 in their

and myc because they are important in deriving or producing induced pluripotential there are must be targets being regulated singularly by these genes. and just for example if we take oc 4 out we can see the pathways highlighted.

the map kinase signaling pathways, 92 genes are regulated by just oc 4 alone. you can see the number of genes in each pathway which is regulated by oc 4. we have situations going on where these genes are regulated in a co-occupied situation, you

need one or two of the other members to induce expression. this is what's shown here. in a number of these genes here, shown here, are genes that require sox 2 and oc4. there are a number of genes involved in differentiation of the germ layer that are

repressed, by these genes in the coordinated fashion. that coordinated repression is due to the ability of these they in concert to regulate the polycomb gene complex which suppresses expression of differentiation genes. there's a ying and yang going

on. there are multiple signalling pathways that regulate the expression of the trinity genes and i'm not going to go through them but some of them are directly controlling the expression of these genes while others are indirectly

controlling the expression of these genes. i've shown here lif, tgf beta-related series of proteins active, in nodal, and the map kinase pathway, some activate expression of the like nanog, others suppress expression and are involved in differentiation.

so what about ips cells? ipc cells utilize some of the same genes important in maintaining embryonic stem cell pluripotency. he won a nobel prize for these, reprogramming somatic cells, fibroblasts in this case. and obviously myc is dangerous,

implicated in onco-genesis, and recently nanog and lin 28 -- there's combining interactions to reprogram adult genes, genes in themselves regulate signaling pathways that are important in driving this process, converting a fibroblast to an ips cells, the three genes here suppress

tgf beta signaling, you're blocking emt. but there's another signaling pathway which stimulates klf 4 which stimulates the mesenchymal to epithelial transition, that's able to take a mesenchymal-like cell and drive it back to a primitive embryonic

epithelial-like cell, you can see the importance of the two embryonic processes, that will become important later. so what is the thesis of the cancer stem cell hypothesis? it was a way of explaining hetrogeneity, i'm going to talk about solid tumors, it has the

same basic tenets. first of all there's two ways you can generate cancer stem you can either generate it in a clonal fashion or stochastic fashion by two tumor cells in the tumor itself have the potential in a random fashion to become a cancer stem cell after

transformation. and that cancer stem cell can symmetrically divide and give rise to a clone, the seed for the tumor from which the other cells in the tumor are generated. now, in 1988, john dick show there's a subpopulation of cells

which has characteristics of embryonic stem cells and they are pluripotent at a rare frequency. these cells he found divided asymmetrically, retained a dna strand and the other strand was given to the daughter cell so the one strand that was

maintained in a symmetric fashion was passed to the daughter cancer stem cell where the other asymmetric strand was given to a cell which then was able to differentiate. so you have a cancer stem cell that already preexists in the natural environment of the tumor

but exists in a small subpopulation. now, this theory has been subsequently modified, these cancer stem cells are not fixed in the tumor from the tumor's inception. they can be derived from any one of the cells in the tumor, and

that's dictated by the niche in which those particular stem cells are -- excuse me, tumor cells exist. you can have a fully differentiated tumor cell or progenitor cell reverting to a cancer stem develop given the appropriate environment or

niche, induced by a transdifferentiation process taking a differentiated cancer cell and driving it back to a more primitive state. this was a thesis that was actually raised and actually proven by bob weinberg at m.i.t. it's really a hybrid thesis of

the classical or stochastic model, and the -- excuse me, the classical stem cell modeel or classic or stochastic model. they have a mixed characteristics, they are both right, it's a more dynamic process than either of these models predicted.

what determines the frequency? that's important. first of all, the cell of origin will dictate the number and frequency of the stem cell. is it derived from a preexisting normal stem cell or more differentiated cell? genetic and epigenetic

modifications that the tumors cells have accumulated during tumor progression dictate how many cells may have the potential to be cancer stem there are contextual signals in the microenvironment sensing, important in determining frequency of cancer stem cells

in a given tumor. what's the niche and characteristics with any given tumor that's going to generate a potential stem cell population? another extremely important factor when you're trying to identify stem cells is the immunological status of the host

of the mouse, in this case, that you're looking at. if you have a genetically altered mouse, the immune stats of the mouse if it's more immunocompromised, you're going to get more. this is what you have to take into account in tumors.

what are properties shared by normal stem cells and cancer stem cells? they undergo asymmetric division and self renew. tissue stem cells, tissue specific stem cells are capable much undergoing self renewal, during the aging process to

maintain various points in the adult differentiated population of cells. cancer stem cells are tumor initiating cells, they can give rise to tumors. they are the ones that have that potential, not the differentiated tumor cells.

that implies a capacity for tumor promotion. stem cells is the differentiated into heterogeneous populations. if you have a tumor with a specific phenotype, you have to have a stem cell that give rise to the same phenotype from the original tumor from which you've

isolated those cells. finally, the normal tissue stem cells and cancer stem cells are regulated by similar intrinsic genetic pathways, i talk about three genes which are important, i'll show you later about the cancer context, and extrinsic signaling pathways, jeep rate--

generated from the tumor or signaling molecules generated by surrounding niche, an environment of the tumor. fibroblasts, mesenchymal stem cells, myeloid-like cells, endothelial cells, all sit around the cancer stem cell and provide a cellular niche and

they are all capable of producing activities or factors which are important in regulating the activity or function of those cancer stem that's shown here. so the normal niche and abnormal niche may be two different environments.

and the normal stem cell obviously if it undergoes mutations or transformation in some way through radiation exposure can give rise to a potential cancer stem cell but progenitor cells can give rise to potential cancer stem cells. they are project any may be able

to do that. fully differentiated tumor cells may undergo the process of emt and generate intrinsic cancer it's a chameleon trying to understand where those populations exist and how can we target them in a given tumor? it's a tricky business.

it's going to take a lot of work. this just shows examples of tumor generated from the primitive stem cell within the tissue or more differentiated cells, like in the hair follicles, you can have the cells being targets of

transformation with the potential of giving rise to tumors and be potential cancer it's a cell in flux. the same exists for breast cancer and the same exists for for hematopoietic malignancy. each of the populations can be respective targets for

now, the cancer stem cell unto itself has some unique properties which makes it a very nasty cell. first of all, there's genomic instability, there's increased dna repair. there's abc transporters which are upregulated, these are

proteins involved in regulating eflux of drugs. the drug transported are upregulated. cells tend to be resistant to chemotherapy and radiation therapy, prone or derived from emt, they are basically

quiescent in nature. and they are resistant to dna repair and dna replication are two different animals, for your clarification. there's a number of properties here that distinguish cancer stem cell from other bulk cells making them resistant to

treatment so if we could target cancer stem cell we may be in a more favorable position to treat metastatic disease. pathways in self renewal are normally involved in maintaining stem cells in normal tissue. these same pathways are deregulated in cancer cells.

here are some examples of tumors from aberrations in the three major signaling pathways which are themselves embryonic signaling pathways, that are due to protubbations in the pathways. genes that are misexpressed give rise to skin cancer, squamos

cell carcinoma, mammary cancer. what about expression of some of these embryonic pluri potential genes? i've given you a list of tumors, there are a number of human cancers where there's an overexpression of these pruripotentiality genes in the

tumors, a good association between overexpression and high tumor grade and prognosis with respect to the expression of these embryonic stem cell genes or nos genes. in human cancers there's also an association of the targets for which those nos genes regulate

their target genes and aggressive tumors, the study conducted by bob weinberg. if you look at the embryonic stem cell module by transcriptome analysis, you can see that they are overexpressed in a number of different types of cancers.

it turns out looking at breast cancer as an example, it's also been seen in liver, lung, prostate and gastric cancers tumors are an activated esc-like module or transcript ome, target genes that are regulated, generally those patients have a more poorer survival potential

and they have a shorter metastatsis-free survival interval. so it's a signature pattern for more aggressive disease. that was a very seminal finding. if we look at breast cancer, what i work on, and particularly embryonic genes reexpressed in

breast cancer, it's affecting a number of women, probably a quarter of a million women each year, 40,000 deaths, risk is one in eight women. risk factors is early menarceh, nulli or late parity, and more interestingly obesity, pre-menopausal women increases

risk of cancer by 70%. again, suggesting that there are environmental factors here that may be important. and finally, 5 to 10% of breast cancers are familial, in large part, but not all, due to mutations in brca1 or brca2. we're looking at 90% of at least

breast cancer being due to somatic mutations or alterations in the environment, not mutation, or epigenetic al operations contributing to the etiology of the disease. we have the human breast, and it is consisting of a series of ducts, connected to the nipple,

and the active component of the end of the ducts is a terminal duct or lobular alveolar unit. the mouse gland is different, with a ductal pattern of terminal end buds that basically are at the end of radiating ducts that grow like a tree. there's hyper branching during

pregnancy and during lactation you get milk production. human breast cancer was originally histologically identified, there's a number of different subtypes that you can identify either in noninvasive or invasive breast cancer.

it says that this is a distinct set of multiple diseases, or it's a cancer arising from must be cell types. that's probably true because we now have a system that was developed by charles perot able to subcategorize human breast cancers based on a transcriptome

analysis. there's claude-in, basal and her2 and luminal. the first is triple negative, her-2 are refractory, extremely deadly and more mesenchymal in appearance. her-2 and luminal are more

epithelial. more differentiated characteristics, arising from differentiated cells along the lineage pathway, the more treatable tumors. so the less aggressive tumors are the ones er-positive and these guys are the ones that are

really the nasty ones that are virtually untreatable in most cases unless it's detected early. the mouse mammary gland undergoes cyclical events post natureally. i wanted to show you this in more detail, the duct-like or

tertiary and secondary ducts, and during pregnancy they develop lobular-like structures that start secreting milk proteins. the stem cell population exists at the growing end of the ducts, and this is a termal end bud, it looks like a bud.

it's actually a mixture of different cell populations, there's a basal population around here, and there's a luminal population around here. now, within the basal population there's a subpopulation of potential stem cells, and these cells are interacting in a very

exquisite fashion with cells win the mammary gland not part of the duct that don't give rise to epithelial cells but there are fibroblasts and lymphoid-like cells and they contribute to the maintenance of these stem cells within the primitive duct of the virgin mouse mammary glands.

you can measure stem cell activity by a reconstitution experiment. you kill the bone marrow, radiate, and go in with a donor population of cells and see if you can reconstitute a functional bone marrow. you can take the gland of a very

young animal, three weeks old, clear the fat pad, remove all of the epithelial cells from the fat pad surgically and you deepithelialize and have the stroma. now you can introduce by transplantation a recipient population and those cells will

reconstitute the mammary gland if they have the potential stem cells or progenitor cells, assessing stem cell activity in the mammary gland. that's been very useful to identify growth factors in genes involved in maintaining stem cell activity.

it turns out that the most primitive stem cell in the mammary gland in the house is basal cell population but needs a supporting luminal cell population. these cells which are generally hormone receptor positive respond to hormones by

maintaining the stem cell in thive differentiated state. you can get abnormal mammary gland development suggesting you can get cancer development by perturbing not only the target stem cell but also maybe a supportive cell that's necessary.

you have potentially two targets here which might be interacting in a cooperative fashion. you'll see how important that is in a few minutes. this is for breast cancer but it's true for all types of cancer. you're using tissues for cell

lines and what you can use is various self surface markers to isolate populations of cells that would be indicative of stem some of these markers have been characterized as marking putative stem cell populations and you can isolate and segregate these populations.

it turns how the you can measure functional activity in vitro and in vivo. in vitro with stem cell activity they will form tumor spheres on nonadhered plastic in serum-free conditions. they exhibit self renewal if you disassociate tumor spheres from

one generation, pass them to a second and third and fourth. that's a property of a stem cell. self renewal. that's in vitro assay. in vivo assay, more stringent, is the ability of these stem cells to produce tumor upon

transplantation. under, this is important, limiting dilution. if it's truly a stem cell, very few are needed to give you a tumor. you can go to very low delusions if it's truly a stem cell or progeninot cell and recapitulate

and that's basically the most stringent assay for defining at least stem cell activity. in the breast markers have been used to isolate stem cells and in the case of the human, you have a luminal progenitor in the mouse there's a luminal

progenitor population and a primitive mammary stem cell this is mouse repopulating units. each species has a district marker, in this case epcam and in this case cd24. actually there are a number of markers which have been

identified that can segregate mouse and human mammary stem cells, some are listed here. some of these same markers have been shown to be useful in isolating breast cancer stem in those stem cell populations, cd44 high, cd24 low, with bob weinberg found was that there

was a high expression of these nos target genes again in breast cancer, and this expression of this nos or embryonic stem cell signature was more prevalent in tumors that had a higher grade, more malignant. okay. it was also more prevalent in

tumors that were generally hormone receptor negative, either er negative or pr negative. and so that was a very interesting and helpful, you know, finding with respect to associating aggressiveness with expression of embryonic genes in

these stem cell populations and correlating them with patients or histological aspect of the if you look at survival, expression of these es-related or nos target genes in several studies were shown to be a poor indicator of survival. and there's an association with

er negative, large tumors and triple negative tumors, tumors of basal cell -- basal-like phenotypes. so again those more mesenchymal tumors are higher in stem cell. tumors, luminal and her-2. if you look at the survival or

nos genes, not their targets, there's an association with overall survival. so if you look at oct 4 or nanog, you can see by themselves or together there's a stratification into poor survival in the tumors that are expression of oct 4 or nanog or

together. not only are the target genes important, it's also possible that the nos genes are expressed in breast tumors and correlates with poor survival and more aggressive phenotype. markers have been used in different types of tumors to

isolate potential cancer stem cells, there are a number of cell surface markers, cd 133, 44, these are drug transporters, one used as a marker to isolate stem cells from different tumors. they are starting to collect a

panel or signature by which we can use to isolate potential stem cell populations. i've mention the emt in the beginning when i was talking about normal embryonic development, how that's important generating these areas, derivatives, and the

processes are utilized in early embryonic development and used in cancer and pathological conditions like wound healing and organ fibrosis. emt is taking an epithelial cell, a polarized cell with a district orientation, adhering to each other through lateral

junctions, tight junctions or adherent junctions, sit on the membrane and through the process of emt they become more mesenchymal, you lose all of these epithelial proteins that are involved in maintaining cell polarity, tight junctions and cell adhesion and acquire a

protein involved in maintaining mesenchymal phenotype. this is regulated by a number of transcriptional repressors, snail and slug and zeb 1 and 2, and they tend to repress expression of these epithelial genes and they induce by acting

as co-stimulators expression of mesenchymal genes and these genes themselves are the target genes for a number of different signaling pathways. they upregulate expression of the genes and the signaling pathways are contributing to the

express of transcriptor factors and driving emt. the reverse situation is also true, signaling pathways like the bmt pathway that actually repressed expression of the mesenchymal proteins and induce expression of the epithelial this is driving mesenchymal to

epithelial transition. how is this important? well, emt is important from for metastasis, if you want it to transit and colonize another organ how do you do that? epithelial cell has basal polarity in a fixed position. mesenchymal cells are more

motile and if you push them through emt and provide the right stimulus they can enter the circulation or lymphatics and then they can go to an end organ and revert back to epithelial phenotype and colonize that organ. this whole process of emt and

met is reversible and is a process important for taking a primary tumor and getting it to a point where it can intravasate and to survive it has to have an appropriate soil and there has to be appropriate factors in the end organ to support growth, some take the cells and convert

them back into epithelial cells. it's a exquisitely controlled process and it's trying to understand the factors that are involved in regulating this and this which respect to the metastatic process itself. now, emt not only is involved in metastatsis and basically a

marker for aggressive phenotype, generally mesenchymal-like cells in a tumor are generally more resistant to drugs, more resistant to radiation therapy, more migratory, more invasive and these properties unto themselves suggest they may be that observation was made by

several laboratories, it turns out if you drive an epithelial cell, a normal cell or a cancer cell through emt they become more stem-like in phenotype and the cells more mesenchymal with a stem-like characteristics in the sense they are able to act as cancer stem cells if you're

looking in the context of cancer or tissue stem cells, that process of emt is highly controlled by factors secreted by tumor cells and the host environment or niche. this is from breast cancer. if you overexpress embryonic

genes, if you overexpress them all the mesenchymal-like genes, slug and snail go up. the epithelial-like genes go down. if you knock the genes out, you can then reepithelialize, overexpressing them will drive them into a more mesenchymal

state or if you take the tumor cells and knock them down, they become more epithelial in characteristic suggesting these are regulating cancer stem cells by their ability to regulate at least one process which is embryonic process emt. breast cancer development uses

embryonic cells at various stages and growth factors and signaling pathways, so you have a normal breast, a transforming event occurring, a hyperplasia, atypical ductal hyper plasia, an early event. some lesions can develop into ductal carcinoma in situ,

d.c.is. this is noninvasive breast the lesions are encapsulated but once the basement membrane integrity is compromised and once the cells become more mesenchymal and invasive and motile all hell breaks loose, there are a variety of host

tissue factors at the stages. in breast cancer sites are lung, brain, liver and bones with various factors being identified in organs that maintain breast cancer cells. they may facilitate foaming to these various organs. in the primary tumor you have

macrophages, and myeloid suppressor cells, from the bone marrow, secreting factors that are generating and fibroblasts themselves, the cancer associated fibroblasts secreting factors that can take a tumor cell and make it more mesenchymal, more deadly, the

capacity to invade, get into lymphatics and bloodstream and colonize an end organ, in this case the lungs, factors that are basically there and they facilitate outgrowth of colonies. you have extracellular protein important in promoting adhesion

of tumor cells in the lung parenchyma, potential targets with lymphoid like cells, myeloid like cells which facilitate this phenotype. how can we treat this disease? well, most treatments to date have been by conventional chemotherapy or radiotherapy,

treating the bulk tumor population, giving a drug that's disrupting dna proliferation. micro tubular disrupting agents, but cancer stem cells are basically kwaiesent. b asically quiescent. we need therapies that target

the stem cell populations. these are the cells that give rise to the metastasis. it's a win-win situation. so what are potential targets for cancer stem cell therapy? i've alluded to some of them. these are all important pathways that are engaged for maintaining

cancer stem cells. if you target specific aspects of these signaling pathways you may have therapeutic for knocking out cancer stem cells and some drugs have been taken to task in a clinical setting. likewise if you use these sorts of drugs that target e

signaling pathways maintaining cancer stem cells with conventional therapy you're attacking the cancer stem cell population and more differentiated population, you hit them simultaneously. that may be the way of the future in this scenario.

obviously you can attack various components within the niche and signaling pathways which are engaged by those niche components, endothelial cells, cells and macrophages, some factors which are secreted by the niche cells may be appropriate targets because they

are necessary to support the cancer stem cell. here is another potential target or targets in the niche. not only can you target cancel stem cell, you should look at host cells in the organ or metastasis for potential targeting to do this.

in fact drug companies have come up with a series of drugs here that can target signaling pathways, they have looked at the markers themselves as potential targets. and they are also looking at microenvironment, factors secreted by the cells and drugs

that can potentially attack those cytokines or lymphokines. and finally driving to a more cell treated by conventional chemotherapy. this is what we're left with. present day oncologists, cutting the tree down but in the process of cutting the tree down in the

branches he may be giving rise to a tumor of a more aggressive phenotype, maybe what they should do is chop the tree down and going after the roots. and the soil, the niche, which surrounds the roots. i think that's the way cancer in the future will probably be

attacked, and i'd like to thank you for your attention. yes? no, no, they don't create hybrids, they are secreting cytokines fostering tumor cells in the primary organ. i don't know if people looked at cell fusion as a potential

mechanism. i don't know if it's fusion per se or diffusible factors or cell-cell contact. i don't know of any studies that looked at, you know, infusion, nuclear infusion, i don't know. let me tell you a colleague of mine, what he's done, he's taken

human breast cancer cells, mouse tumor cells, and he's taken and mixed those cancer stem cells with normal mouse mammary epithelial cells, okay? and what he's done is gone back and labeled tumor cells, genetically labeled them. all the tumor cells now convert

to a normal mammary gland. irregardless whether they are a mouse or human, what does that say? there's something in the niche] that's able to reprogram across species, tumor cells. what are those factors? what are those cells that are

interacting with the tumor cell? are they resident cells in the mammary gland? are they lymphoid invading cells? are they stromal cells in the fat pad, or is it a collection of characters? yes?

it will keep people busy for a long time. but it offers the potential of new therapeutic targets. okay? out of the box. what? ah, there is lies the rug, as shakespeare instead, you have to

kill the tumor without the organ from which the tumor arising, a delicate process defying what factors are expressed in the tumor niche versus the normal niche. and how does the tumor stem cell differ from the normal stem cell?

it's a dialogue here that's going on. cancers basically are a dialogue. it's an antisocial disease. cell times are talking together in the wrong way, it could be disastrous, that's what cancer is.

but it's going to be tough.