Avastin Ovarian Cancer

>> we're going to get started. thursday is thanksgiving day, so i wish you all a happy holiday. today our first speaker is jung wei, ph.d. at baylor college of medicine, did postdoctoral training at the national human genome research institute with paul meltzer, paul came to the

genetics branch and he joined a lab, they are in the oncogenomic section of the ccr, the title today the application of genomics to identify diagnostic biomarkers, drivers and therapeutic targets for pediatric cancers. >> thank you, terry, for the

nice introduction. thank you for coming in the cold weather to attend to this lecture, really it's my honor to tell you how we use genomics to study the pediatric cancers. this is the outline for today's lecture, the first i will introduce you to success and

challenges of treating pediatric cancers. and the second part i will tell you what is the genomics. the third is this technologyy using next generation sequencing, two examples how to use next generation sequencing technology for the particle use

in terms of diagnosis and also identification of molecular target. lastly i want to touch a little bit on the precision therapy, which is the new wave of clinical care for pediatric cancer. this slide shows in the child

the cancer therapy, the modern therapy, the survival rate is increased in all kinds of pediatric cancer. here shows leukemia, lymphoma, the wilms tumor. with 90 you can see most cancer survival rate is in 80%, even 90%, such a successful story for

pediatric oncology research. it's a successful story. and this is the same graph that shows the mortality rate keeps declining from the '70s to the '90s. however, in the recent several decades you see this decrease rate plateaued, okay?

and that means probably this kind of approach using empirical approach reached its maximum effectiveness. in some leukemia, the survival rate as i've shown it's pretty high, but the high rates of neuroblastoma, even in the '90s it's still low.

some tumors, for example tumors in the brain stem, is very low. for example, ewing sarcoma, localized disease, you can see survival rate is very high, but in the distant metastasis disease the survival rate is very dismal. the same trend we'll see in the

rhabdomyosarcoma, if you have a metastatic disease the survival rate is very low. so this fact that the success for improvement of the successful rate for the childhood cancer and a certain tumor especially those metastatic tumor, the survival

rate for those patients is still very, very low. this slide shows you hoe important genomics is. let's stand talk to talk about a phenotype and genotype, in this slide this is dramatic change of morphology from the caterpillar to this butterfly.

you have the same -- the organism sharing the same genome but it looks so different. that's the dramatic consequence of gene expression of the genome. it's the same genome but it has a different expression pattern, and the result into different

proteome, tissue and physiology, the same genome results in such a difference. the same in humans too, we all come from the single fertilized egg but in our body this is so many kind of tissues. the skin, the epithelial, the gland, so on and so forth.

i don't need to bore you with this but the same kind of idea, the same genome, and you have a different expression pattern and different proteome, different tissue and different physiology. so this is early '90s study of the human genome, the human genome project is a landmark to

decode the human genome, so that now we have a map to really know what a human genome looks like, and it's really this project that greatly facilitates the advance of the genomic research. essential biology. the genetic information flws from dna, this is chromosome,

dna, to transcribe into the mrna, messenger rna, into the protein, so the central dogma of the biology, the information flows this direction, and of course is a regulatory feedback approaching we know that it can regulate the dna and mrna level, recently we know there's

some regulatory rna, microrna, it can also regulate the transcription from the dna to the messenger rna, also interfere with the translation, the translations themselves, they can self regulate. it's very complicated system, but this is the current

understanding of how this genetic information flows. and this is the number of this molecules, the dna we know we have three billion base pair of dna in every cell of all humans, human body. this three billion base code dna encode 25 to 30,000 genes,

alternative splicing isoforms there's 150,000 species of messenger rna. and there's also about a thousand microrna, and over 10,000 encoding rna, the regulatory elements, regulatory, the process we just talked about.

and the protein, this is our explanation, we think it's about half million different kind of proteins resulting from alternative splicing or the variants. this genetic information flow resulted in protein and this protein is really the building

molecule for many phenotypes, right? in cancer, to really -- this is the protein that caused the specific characteristic of the cancer, and also we can use this protein, the phenotype, to diagnose this cancer and also to affect the response of the

treatment, kinds of treatment. and study revealed 80% of the human genome, functional, before the encode study we don't know. we know these three billion base pair of dna, it's only 1% or 2% is coding dna, the rest we call it so-called junk dna, people call it.

now the encode shows about 80% of the genome actually they have been transcribed and they have function, major function is the regulatory function, to ensure the right time, the right place, and the right moment the protein -- the gene transcribed, the right isoform of protein is

expressed. so in the genomic research in the cancer, the key is to use genomics to identify the biomarker, the driver and the the perfect scenario to identify this is we can identify biomarker, which is molecule that we can measure objectively,

okay, for example just like the blood measurement, the red blood cells, numbers, it's all biomarker, you can objectively measure it. so if you can identify biomarker, and hopefully the biomarker is specific to cancer, so that's the reason we call it

the driver, which is specific to the cancer progression or to formation of the cancer, so this biomarker cannot only as a marker to indicate the cancer, also they can -- at the biology lab, this biomarker also drives the formation or progression of the cancer.

if we can identify such drivers, then we can device therapeutic methods to interfere with this driver, so this will be this therapeutic target. however, in some case the biomarker can be the therapeutic target directly, not necessarily the driver.

for example, if there's a cell surface molecule that is specific to the cancer cell, but not necessarily to drive the cell, specific enough, we can use some antibody therapy or other things to target this molecule in order to target a so that's really sometimes the

biomarker not necessary to be the driver but it can be also the therapeutic targets. in any case, all this kind of scenario requires this kind of biomarker or driver or therapeutic target had to be low enough, low expression in normal tissue, early expression in

cancer, right? that makes sense, right? that's the genomic research which tries to target those kind of molecules, low expression or no expression in normal tissues. that's the goals of genomic research in it the cancer field. so here i want to introduce you

to this technology which is called next-generation sequencing, also called massively parallel sequencing. you will see why it's called the idea is in the old time this one kind of genetic truth, if you have a human genome, it's actually a lot, it's very hard

to sequence in the old times. so what they do is they fragment into different fragmentation, and then they clone into the vectors and sequence piece by piece, right? that's traditionally sequencing of the human genome. in these kind of methods the

first of human genome project spent about 13 years, 13 billion base pairs, and about 10 years to finish one human genome using that kind of method. but during that period of time they used another approach, using the human genome dna to randomly fragment it and

randomly take those clones to the certain sequencing primers and randomly pick clones to sequence them. so that way you can eventually create a (indiscernible), because they overlap, build up the long fragments, eventually you come back to the human

so this idea next-generation sequencing is similar. genomic dna here, fragment the dna, you do a size selection, and then you put a sequencing adapter and two end of the fragments, and then you amplify it, amplify this and sequence next-generation sequencer.

haven't eventually you have the reference genome, because at this point we have the human genome sequenced. we know what the genome sequence looks like. you can line up with the reference genome, and from all

this experimental sequencing you can see what exactly sequencing is original sample here. so that's the idea. and this is a real picture from so each little square here you can blow up is this picture. each little dot is one sequencing, okay?

it's equivalent to run sanger sequencing. each spot is one sanger sequencing, and there's a hundred million spots in the flow cell. this is only one -- this is only one little spot of the whole the whole flow cell has hundreds

or thousands of this field. so you let dna flow through this flow cell and you sequence it, you take those pictures and then you sequence by synthesize. what kind of genomic information you can get from this sequencer? this shows at the dna level you can detect that the point of

mutation, for example this is the reference genome, if you have a resident genome at this position, letter a, your letter is c, then you know that your sample has a mutation, a to c point of mutation. and in this case, you -- it's indel.

this also sequence is missing in all the reads in your sample. this is an indel. also if there's no in one region, it is absent,'s it's a homozygous deletion, there's nothing there. if you do compare with normal samples, you will see the

coverage of particular segment is reduced, then you can know hemizygous deleti because of counts of coverage. then of course you can detect the gain, or amplification, or if you detect the junction, that is not natural in the original human genome, but two sequence

coming from different chromosome, then you know there's a translocation. there's a break point and a cause of translocation, okay? so another thing is you can also see those coming from the experiment, map back to the human genome, map to some

non-human sequencing, for example, virus or bacteria, then you know in your sample there's a pathogen present. this is at the dna level. and this slide shows you the rna level, you can do the similar thing. first, with the rna

transcriptome sequencing, what you can do is just like a microarray, you can by the count of the reads from the gene you can kind of know how much a gene is expressed. all right? so this is called digital gene expression, by count, not by the

analog signals as seen in the microarray experiment. and the thing we have base pair resolution by expensing that it will know that expressed mutation, if the mutation is expressed or not expressed we can immediately tell. and since the rna has to go

through the splicing to become a mature mrna, and then using the sequencing we can detect alternative splicing event because the junction we can see the junction right at the boundary. we know exactly which transcript coming from which kind of

splicing pattern. another very important feature of the rna sequencing is to detect the expressed fusion transcript. remember the fusion gene? if it's lending to a coding region they can fuse together and get a transcribed fusion

gene, and then we can detect it readily. this is a very important feature. and also there's other things like rna editing, novel transcript, this technology can really study all this kind of events.

what's the good of the next generation sequencing technology? this slide shows compared to the old high throughput method like the array-based technology, the next generation sequencing technology has no need to prepare clones for dna

fragments, right? you can have all the dna species detected in the sample. and then there's no need of prior knowledge for probe design, which is absolutely required for microarray. microarray you can only detect the things that you kind of

know. if it's new transcript or mutate like a deleted gene, it's really hard to detect because you have to design special probe to detect the specific event. and another thing you can detect, able to detect the balance of genomic structure

change. for example, if there's a balanced translocation, this kind of event is almost impossible to detect in the array-based format, because there's no net gain or loss of genetic materials. the array cannot detect that.

the next generation sequencing you can detect that. and of course the parallel sequencing is giving the base pair resolution and the very massive throughput, and can give you extra information in terms of mutation. and here i call it cheaper and

faster per genome, okay. experiment-wise it's still pretty expensive. for example, if you nowadays want to use -- sequence a whole human genome with the technology it still cost ten thousand dollars to sequence that, but compared to the first

genome, the first genome we spend like a billion dollars and sequenced one, right? it's much cheaper and faster, the first human genome takes something like ten years, but nowadays you can sequence one human genome at 40x or 50x almost overnight you can do

that, okay? this technology is very powerful technology. this slide just shows we can get all different kind of information, for example i showed you from the dna we can do the whole genome. this is variation of genomic

genome partition you can use probe to only capture the part of genome you want to sequence. for example, the coding region, okay? you can design probe to only target coding region. so you do not extend a lot of sequencing ability to sequence

intergenic region. you can study methylation and chip-seq and transcription factor or in cancer, the binding area, so you can get all different kind of information from sequencing dna using different variations, various techniques.

the rna level, you can, as i showed you, you can really sequence messenger rna, non-coding rna including microrna. all these things you can get from the same platform, okay? so one platform, one generation, next generation sequencer, can

generate all these kinds of different information for your sample. so this is -- i've already gone through this so i'm not going to go one by one. again, the goal of this is we can discover the biomarkers for that diagnostic or prognostic

information, and then we can understand the biology to identify what is the driver driving the tumor initiation or progression, or identify i will give some example later. and this slide shows one example of the application using gene expression profile to diagnose

and actually this data is coming from the microarray, the early microarray experiment, using metrics micro away, we did the whole genome expression profiling of the tumor, and this is -- this gray dot represents a tumor -- this tumor, the clinician has a question about

this tumor. they do not know if -- original diagnosed was wilms tumor, a kidney tumor, but in the clinical process they are not sure because it doesn't look like a typical wilms tumor. they come to ask us what kind of tumor it is.

we did the general expression profile, from this pca you can see this is -- instead of cluster with this wilms tumor showing yellow, the cluster was neuroblastoma tumor. we changed the diagnosis from wilms tumor to neuroblastoma, the patient was switched to

neuroblastoma treatment including stem cell transplant and this patient is doing well after one year of the diagnosis. this is really a good example using molecular diagnosis, if the genomic tools to diagnose the tumor, it's sometimes much more accurate than the -- like a

regular histopathology diagnosis. this is another example shows that use next generation sequences of transcriptome data, distinguish subtype of rhabdomyosarcoma, so this red represents here is a fusion-position tumor, which

those tumors are possessing typical fusion molecules that grab those tumors. and this category tumors, it's all fusion active, they do not drive by the fusion event, instead they are driven by mutations, in ras pathway, other kinase pathway.

you can see using next generation sequencing, sequencing rna sample, you can readily distinguish these two pep sides. and then another example to do the diagnosis is to detect fusion, fusion gene. fusion gene is very important

diagnosis marker for the pediatric cancer, many pediatric cancer for example rhabdomyosarcoma to show you the fusion positive, they have this pax3-fox01 fusion, or this is ewing sarcoma have this ews-fl i1. how many people are familiar

with this graph? okay. good. let me just for those people who do not know the graph, just s graph actually is theexplain to you briefly. circle part, it has multiple tracks from outside to inside. okay, this track is to represent

human genome. these are chromoes 1, 2, 3, 4, so on, all the way to 22, and then x, y. and here the second track here is represented the copy number, you can see this is probably represented two copies, and if higher that means there's a

duplication of the chromosome, for example here probably you have higher copy number, maybe again of the whole chromosome, of the five chromosome 5 in this tumor, and if i'm here, you can see this probably is the boy, because it only has one x chromosome, and in this case the

rhabdomyosarcoma case you can see this chromosome has a deletion, ocean? so one copy deletion. so you have chromosome 11 at the beginning, 11p has a deletion, you can see the copy number is lower. and the inner track here those

the heterozygosity and this inner track is the variance track to show that the variation of the sequence from the reference genome, the human genome, okay? the utility using whole genome sequencing immediately you can see what kind of tumor it is

because the characteristics of translocation, okay? it's very easy to detect with this technology. and also the same kind we can detect the novel fusion, this is rhabdomyosarcoma, and we detect the fusion in the dna, pax 3 instead of fox01, the typical

partner of pax3, this case fused with the gene which is epigenetic genes, and you can see this region has a massive rearrangement in the sample, and this is rna sequencing, immediately we can see this is the fusion junction to show the fusion, and you know this gene

fusion is not only fused at the dna level, but also expressed in the rna. so probably this is the function of fusion. and it is two cases when we do this sequencing of rhabdomyosarcoma, at that time we did the sample from the

tissue network bank, and the two cases very interesting, use whole genome sequencing we detect originally histology of sarcoma, differentiated rhabdo sarcoma, the alk, then the tumor bank, they review the histology and they said most -- this is like hematological malignancy,

probably one type of lymphoma, also reported as this kind of alk-npm1 fusion in the literature, clearly this is a sample messup or misdiagnosis at the beginning for this case. a second case is similar, we detect ret-nco4, either the

diagnosis was wrong or sample has been mislabeled. so from these two cases we can see using this kind of technology it's very easy to pinpoint what exactly is the diagnosis of a tumor. and this is a study i want to show you how to identify the

novel targets for pediatric and this study was conducted for case study for a patient with neuroblastoma, the slide shows neuroblastoma is aggressive childhood cancer, usually happens early, the first five years of the life, so 90% of patients neuroblastoma happens

at the -- before age of 5. they have very diverse outcomes. i will show you more data for that. and the survival for the patient with high risk neuroblastoma is less than 50%, despite intensive multi-modal therapy including chemotherapy, radiation, bone

marrow transplant, very aggressive therapy. so this therapy can result very toxic cytotoxic therapy can result in high morbidity for the patient. high mortality rate, especially with the patient -- especially patients with metastatic tumor.

so novel targeted therapies are highly desirable. and this is what stage 4 neuroblastoma looks like. this is a bone scan. you can see the tumor is spread everywhere in the bone. this is a big tumor in the -- in the adrenal gland.

this is tumor taking adrenal grand, a viable tumor cell, extensive necrosis, you can see the histology, the tumor cells. usually those kinds of patients, their survival rate is less than 50%. specific stage 4 disease for neuroblastoma, stage 4s, also

the patient, you can see, has very extensive metastasis. this skin shows a lot of tumors, nodules in the patient. same thing, tag the tumor cells. however, this kind of patient usually the survival rate is 90% because those usually happen in the patients younger than 1 1/2

years old, and as long as you can keep the patient alive, and the tumor amazingly they spontaneously regress, it just goes away, okay? so that's the interesting about the neuroblastoma, the clinical outcome is very diverse. and this one shows the high risk

of neuroblastoma, it's still a challenge for the treatment. the immune therapy treatment survival is about 50%. so our section, the mission of oncogenomic section is apply this to study the biology, to identify the novel target, and the biomarkers, and hopefully

translate such findings to the clinic. this is neuroblastoma genome, the large study by the target group, which is like pediatric genomic study for the neuroblastoma, and this is the landscape global view of this type of tumors.

you can see there's a lot of amyploidy, and very extensive genomic rearrangement for this tumor. however, if you look at the mutation landscape, only a small portion of samples having no genomic mutation including alk, pdp and 11, so on and so forth,

also the germline mutation has barred1 and dna repair mutations. look at this, the majority of the sample actually we cannot detect any mutations. and as a matter of fact, this tumor has the lowest somatic mutation burden, okay?

this is bert vogelstein's review paper published a couple years ago, and you can see mostly adult cancer in this space, this is pediatric cancer, you can see it's all on the very low end of the mutation, this is neuroblastoma. so this is a case study.

we did it. we performed the next generation sequencing several years ago, so when we first did the -- the first sample we got is from this diagnosis sample which is from bone marrow biopsy diagnosis, the tumor content for this is about 80%.

and then this patient went through the chemotherapy, eventually she wept through the surgery to take out the primary tumor, as shown before, and this is marginal, viable tumor cells, we took for the second samples as the primary tumor. and this patien after the

surgery went through three years of different kinds of therapy, and unfortunately she eventually succumbed to the disease, and necropsy we took one metastasis tumor from the liver, her liver, as further sample. the idea is to use the samples to really use the whole genome

and the transcript sequence to identify if we can detect any mutations that we can use as a target to treat the patient. they examined the design as following, first sequenced the index metastasis, met2 with germline dna, used the whole germline sequencing, identified

variants, and then we look at -- we use the orthogonal resequencing, making sure it's real mutation, not sequencing artifact. and from there, we went back, this is the index sample, we went back to original -- the sample at diagnosis to see how

many dose mutations are at the beginning sample. and also was this primary tumors we have different sections, because that tumor is pretty big, we want to see if this is mutation is in all kinds of different parts of this tumor. we also integrated transcriptome

sequencing to see are those mutations expressed in the transcriptome. this tumor has very extensive genetic genomic rearrangement. you can see 1t, 3, 11, 14, so on and so forth, again the copy number and a bunch of translocation happening in this

and at some places for example it's too example, chromosome 4, 13, you see very small genomic regions, this is only like a couple, but you have this expensive rearrangement that indicates the phenomenon called the genome shattering. so since catastrophic event

happened to this genome and the genome put back into this kind of really rearrangement. the gene we're interested, in the literature has reported for this kind of high risk of patient, especially older patients, this patient is very unusual patient, when she was

diagnosis of neuroblastoma, 90 years old already, and for this patient, older patient in the literature reported atrx genes mutationed at high frequency. we looked at this gene by looking at the coverage around this region, around this gene, and immediately you see there's

a deletion for this exome from 10 to 12, there's a deletion. so then we designed a genomic primer to verify the deletion, if there's no deletion the distance is too far. you won't have the -- the pcr won't work. if the deletion outnumbers the

junction, you can result pcr product, that's what exactly we saw in the diagnosis tumor, in the primary tumor and in the autopsy tumor we can all detect this junction. but in the normal liver, normal skin, we can detect that and sequence verify the finding from

the whole genome sequencing. this table shows you, you're not supposed to read those, the reason i put here is to show you this is red each row here represents one mutation, okay? his is red portion, representing mutation that's shared in all the samples.

what's black is only detected in this met2 but not the other samples. that means that there's an accumulation, rapid accumulation of de novo mutation, somatic mutation, probably just the heterogeneity of the tumor, or it's due to the intensive

therapy, so at this time we really don't know what happened, but this is very alarming how much mutation, de novo mutation can happen. also there's a mutation we look at expression, this panel a, this on the left-hand side depict shared mutation.

upper panel sh the d, and the lower panelexpression of the genes that is shows mutant allele frequency in the sequencing data. so we can see that this gene lpr1 is highest expressed mutated gene, mutation rate is pretty high, more than 50% in the sample, indicating this is

important mutation. and interestingly, if you look at the mutation that is unique in the met2, the black column i show you in the previous slide, you see the expression is only detected in the met2, not in the previous -- not in the other two samples that's consistent with

dna data. what is this lpr1? g protein-coupled receptor, it's important for the neuron development and also it's important for the cell migration and the differentiation, also survival. and you can see this lpr1 is on

the top of many pathways for the cell. and this particular mutation is at the second intracellular domain, right at the junction of this intracellular domain with this transmembrane domain, predicted to be deleterious for this protein function.

and so we clone this mutated gene, and with the wild-type we expressed these genes into the fibro blast cells and performed the gene expression profiles to see what kind of gene profile list of the mutant, and interestingly, the gene signature of this -- so the

paptotaxis for the cell mobility, the cell mobility genes showed up, another interesting category is this row a pathway is upregulated. we look at cell growth characteristics, 1%, 10%, wildtype mutant does not have

difference. we can see those -- either we use the fbs or lpa, the mutant form has elevated mobility shift. a rho pathway, gtp, the mutation has elevated response to the ligand, than the wild-type.

if we put this system into using another reporter downstream of the rho pathway, luciferase reporter, you can see the mutant has more signaling function than the wild type would show in blue. this data indicates that this mutation really results in

enhanced signaling through the rho-rock pathway. this shows the mutant form has much faster cell mobility than the wild type, but if we use this inhibitor, low kinase inhibitor, we can stop this. in the literature, another study published in 2012 shows the

neuroblastoma in that study is detected, this rho-rock pathway enriched with the mutation, and this is the diagram from that paper to show that the balance of the gtp, rho activator, many members mutated in that study. and also the gef, which is the opposite of the activator, the

inhibitor of the rack signaling, mutated. so this data is consistent. we think probably activation was associated with defect of the genes in the neuroblastoma, so the cells cannot go through the tunnel of differentiation. here is the summary of the

study, the whole genome sequencing showed massive chromosome alteration together with the small set of somatically expressed in the autopsy samples reveal rapid accumulation, that has implication because it might be have resistance coming from this

mutation profile. this dna and rna sequencing enables us to identify the cell motility driver mutation of the lpar1 gene suggesting this approach can be used to use in the precision therapy by identifying the target for each individual patient.

important with high risk we can use this approach to identify the target for each individual patient, so that we can use precision therapy for the future pediatric trials. currently, all these patients are coming in, we do not know which patient has what kind of

mutation, so they are given the same kind of treatment, the treatment depends on their risk of stratification. if it's all high risk, for example here, this is all metastatic disease, they all receive the same treatment. but we can use genomic

biomarkers to really saturate them into different pools of patients, their response to current therapy well, so they don't need to be giving extra therapy. they can give just the conventional therapy, standard therapy, so they will do well.

but for those poor, the patient with poor signature, will probably need to identify what exactly mutation they have, so that we can put them into a driver mutation. so the driver mutation can be amplification, mutation, translocation, overexpression of

gene, alternative spliced genes. so that's the idea. so under this precision therapy idea, currently we're opening protocol to collecting the tissues for all the patients coming into nih and we can perform different genomic study. this will enable us to

understand how useful this method can identify the target in the future. in parallel, there's an effort in the ccr, we're going to use the next-generation sequencing to profile the patient, especially the high risk patient, and so can inform the

clinical management of the patient, so the patient comes to ccr, we can load into this program, and then we can do the germline somatic sequencing, and same time we can acquire the research data as show you in the previous protocol, all this information we can return to the

clinic for guided therapy. also we can learn from that, to go through this cycle so that we can understand the disease better. so in conclusion hopefully i show you that integrated analysis of the cancer genome identified biological relevant

diagnostic, prognostic biomarkers, and novel targets for the therapy. and the powerful tools can help the completed genomic portrait of pediatric cancer at the base level. this will lead to identification of the key drivers and will

enable development of the future novel therapies and precision and this just oncogenomic section, javed khan is chief of the section. we have very talented and hard-working group, in the biological part or the bioinformatic part, because

there's so many data for the studies, that this studies closely to make the study possible. so thank you for your time. [applause] i don't know if we have time for some questions. >> (inaudible).

>> yeah. >> yeah, so because we have a time line of the samples, right, those mutations we do not detect t the very original diagnosis sample, we don't see it in the primary tumor either. they are only at very end of the disease course, so that's the

reason we see the de novo. we sequence a thousand-fold and don't see any evidence of the mutation, yeah. yes? >> uh-huh, yep. >> many cases they do, especially the pediatric cancer, the one i show you, this is

transcribed as fusion transcript, transcription factors that can drive the gene. >> thank you, jun. >> thank you very much. >> so some of the people in the traco course signed up to visit the tumor boards, and at the

tumor boards they go through case reports for various patients. today we bring the case reports to you, so we have dr. olaku in the division of cancer treatment and diagnosis, he's in the office of cancer complementary and alternative medicine.

>> good afternoon, everybody. thank you for giving me the opportunity to present case report. dr. wei presented presented twocase reports and makes my life easier, but i hope at the end of this short lecture hope to achieve describing history of

case reports, recognize potential roles of case reports, describe case reports, give you some examples of case reports, and outline pertinent information for a good case this is one definition of case report, and it's case report is a formal summary of the unique

patient and his or her illness, including the presenting signs and symptoms, diagnostic studies, treatment course and outcome. case reports started a long time ago, and some of the oldest case reports were probably from text from egyptian papyrus, 1600 b.c

it's possible these were copies from previous century, among this 48 cases discussing injuries or disorders of the head and upper torso. you have hippocrates, and hippocratic doctors or physicians, another time they looked at retrospective accounts

emphasizing clinically relevant findings. some hallmarks of hippocratic case reports they focus on findings and observation of the course. both mental and physical findings, however the patient version of the complaints was

for the most part absent. another name is claudius gallen, who introduced conversational tone, placed himself in the first person in the case reports, describing his working day, his doubts, tentative diagnosis, and his interactions with other physicians as the

disease unfolds. it appeared in the middle ages that western medicine was dominant. however, at the same time, islamic medicine seemed to prosper, and the case reports of the time had hippocratic and gallen as concerns astute

clinical observations, the 17th and 18th report gallen was present, more emphasis on the patients' subjective experiences. physicians tried to put drama in in by delaying the moment of diagnosis or outcome of the

story to heighten tension, narrative tension and degree of physician involved. for example, you can see here said a girl 3 years old who remained a quarter of an hour underwater without drowning, case report in philosophical translation in 1739.

19th century dealt less with patients' accounts, focus more on clinical findings. at this time they introduced sections into the case reports for example the demographic details of the patient and outline of clinical cause of events so they used mal

terminology at the time. in cancer patients, there has been a long history of case reports, case reports the first line of evidence where everything begins. as mentioned earlier, some of these were from the ancient egyptian papyrus around 1600

b.c., first recorded in incurable tumors of the breast. several roles case reports played, one of the rules is the recognition and description of new disease, for example in 1999 west nile encephalitis was recognized in the united states in new york when they had cases

of west nile encephalitis, as a result of looking at the cases and the reservoir for this virus enclosed, they were able to come up with methods to reduce, they had to spray to reduce incidents at the time. another role case reports play is detection of drug side

effects, this could be adverse or beneficial. quite a significant percentage of drug retractions are usually due to case reports. people report that they found side effects of that would lead to the body responsible in this country the f.d.a. to retract

the drug. another example beneficial side effect, the drug that was designed by pfizer, in england, the drug was initially designed to treat hypertension, and angina pectoris. the drug did not have effect on however, they found a side

effect that most specific of the men had sustained erection, that was what led to the development of viagra, it came out from another effect of a drug meant to create hypertension. and between depression and smoking cessation, this led to using antidepressant drugs to

treat nicotine withdrawal symptoms because of this association, this came out of case reports. case reports can also be useful in the study of mechanism of disease, for example in this case you have the maternally inherited diabetes associated

with deafness. borrowing a phrase, mitochondrial dna present in all sites, this disease can only be transferred maternally. case reports are important in medical education and audit. if you're familiar with the new england journal of medicine,

almost every week they have a case report. one of the ways i became very (indiscernible) this year when i was in full time clinical practice i learned a lot through it's also important to look at their practice to see what they are doing right and what they

may not be doing so well. it's also important in recognition of rare manifestation of disease, as we go through the cases you'll see examples. the case report is the impact on health policy, if you remember last year when they had the

ebola cases, the first patient in texas, that led to cdc changing their policy in the way patients manage or the way hospitals have to deal with cases like that, or suspicious cases. so case reports are useful in describing new disease or

melanoma was described by hippocrates, and rufus of ephesus reported cases, two of which we know today at hodgkin's lymphoma. 1957, dennis burkitt described a tumor in the jaw of african children, today described at burkitt's lymphoma.

1990, farcet and colleagues described who patients with a new type of lymphoma called hepatosplennic lymphoma. this is a case report, 57-year-old lady diagnosed with skin melanoma of abdominal wall, surgical treatment, in october diagnosed with triple-negative

invasive cancer, mastectomy followed by irradiation, and ipsilateral axillary, and relapse in the brain and bones in may, treated with biphospho-natures, brain metastases treated using stereo tactic radiosurgery,

morphological immunohistochemical determined this was not from the breast. she had five cycles of temozolomide with partial response, a second radiosurgical treatment in november of 2013 for progression of brain lesion. vemurafenibed administered in

2013, grade iii rash third week of therapy, grade 1 back pain, rash diminished after discontinue taigs of vemurafenib. however some popular elements remained at previously irradiated area of mastectomy scar.

the dose was restarted. she received autologous dendritic cell vaccine. after eight weeks she had no evidence of progression of melanoma bone metastasis on ct or mri scan, wever the lesion at the mastectomy scar progressed.

pathology revealed positive er, pr, her-2 negative for all markers, molecular testing demonstrated wild type braf, kras and n.r.a. nras. developed at the level of the third rib, analysis was similar to previously excised breast

lesions. patient passed away may 2014 following progression of metastatic lesions. this story not too clear but this shows the outline when she developed the rash, and then this slide, the lines here was when the initial rash she had,

and then when it extended as the result of treatment, when it was withdrawn this is the solid line showed where it was redrawn, the dotted line how it had improved over a period of time. this is just the histology, and the mammoglobulin, progesterone, estrogen and her-2.

a lady had nasopharyngeal carcinoma, treated with cisplatin, disease free for 18 months, subsequently developed back pain and identified lesions, relapsed metastatic nasopharyngeal carcinoma. she had palliative radiation to the sacral lesion and began

systemic therapy with paclitaxel, after that second line of therapy using gemcitabine, it was controlled, this lasted for 8 months unfortunately. she initially received the chemotherapy, changed to denisomub.

this progressed, she had a third line of treatment with methotrexate, occurred at three months, treated with controlled for four months, developed a new mass in the left breast. mammography identified a mass at the 10:00 position of the left breast b-ras category 5.

we can see the mass there. pet-ct identified the new left breast mass, suggestive of a primary breast tumor. a biopsy was done, consistent with primary breast cancer. immunohistochemistry revealed no estrogen or progesterone or her-2 amplification.

they compared biopsies. breast tumor was similar to the original nasopharyngeal tumor, the epstein-barr virus positive, the diagnosis was changed. at the time of the case report the patient was doing well and began the fifth line of chemotherapy.

this is the histology of the breast, nasopharyngeal tumor. the six month old girl was seen by the pediatrician with three week history of bloody streaks on diaper, presented to the e.r., no history of trauma, parents found a piece of tissue on her diaper.

in addition she also had a heart disorder, being followed by the cardiologist. all the clinical examination was normal except for persistent vaginal bleeding. blood work within normal minute, apart from alpha-f erictoprotein that was high, she had abdominal

pelvis ultrasound which showed a midline mass measuring 3.7 by 30.0 by 3.3 sent meeter in the vaginal region, she had biopsy, large mass with cluster of grapes arising from the vagina. mri of the abdomen and pelvis showed mass, there was no extravaginal extension, but

there was displacement of cervix, uterus and mass effect on the bladder and rectum. bone scan was negati. biopsy revealed a tumor characterized by highly atypical cells with reticular bodies. findings were quest entrepreneur segment with yolk sac tumor.

you can see the tumor here, here on mri, there, and there. she had surgical resection, at surgery tumor was resected completely, no tumor identified postoperatively with cystoscopy, biopsy consistent with yolk sac

tumor, margins involved, negative for vascular perineural invasion. the patient received three cycles of cisplatin, per protocol. afp decreased, and before starting chemotherapy. decrease the after they cycles

of chemotherapy, after three cycles remained at 22. mri showed no evidence of residual tumor. given the abnormality of the tumor, decision was made to give three cycles, after two months, after completion, the afp came down to 13 ng/ml, beta hcg

remained normal, she had no patient continues to do well clinically with no evidence of recurrence more than two years after the completion of therapy. another case, a 37-year-old lady, sixth pregnancy, three deliveries, two miscarriages, went for routine you will that

sound scan at 16 weeks. picked up 16 weeks single fetus, bilateral asymptomatic multi-locular solid cystic masses. the cysts were not vascular. she did not have medical problems or past fily history of endometrial, ovarian,

colorectal or breast cancer. you will that sound scan, you can see cysts on the ovaries. ca 125 was high. ca 125 was high with range of 1015-1025 and they had to have a team to manage the patient. she had exploratory laparotomy

she had two omental nodes measuring 6 centimeters, ovaries removed, part of omentum removed, palpation of the pelvic and lymph nodes were negative, postoperative course was unremarkable. however, consistent with small

cell ovarian cancer stage 3 in both ovaries and the omental nodules. there's the picture. this is the omentum. the tumor cell within the fingers of the surgeon there. three cycles of adjuvant chemotherapy with carboplatin,

ca 125 levels dropped, pregnancy uneventful. she had exploratory exploratory laparatomy, cesarean section, no evidence of residual disease after surgery, no tumor in the uterus, with the section of omentum showing area of fibrosis and foreign body

reaction, female infant 2900 grams, placenta normal, after 23 months doing well with no evidence of disease. now rare cases, it's very difficult to put a indicates series together because there are very few cases available. what does the clinician have to

do is rely on the evidence of few cases to try and make sense when they come across a patient with similar disease. this is a 64-year-old male who presented to his primary care physician with a history of non-specific fatigue, phyl examination unremarkable.

laboratory studies revealed profound anemia, transfusion six units, suspicion of intraabdominal bleeding, they did a colonoscopy that was a large ulcerated mass was shown along the greater curvature of the stomach. and the biopsy demonstrated nest

of cells with large nuclei, expanding into the lamina propria, positive for sox10, ses will 100 malignant melanoma, did not find any cutaneous lesions. they did a pet scan which was able to identify the tumor, surgical exploration showed a mass, excised, with lymph node

dissection. that's the pet scan. on ct scan, the tumor there. pathological assessment of the surgical specimen revealed a large melanoma, the tumor invaded but did not invade the visceral peritoneum, lymphovascular penetration was

evident. that's the tumor, from the specimen. again, they confirmed the histological assessment with possibility of the mart-1 and s100. they decided they were not going to give adjuvant therapy, 18

months later patient presented with small bowel metastases not amenable to resection, unfortunately he passed away with diffuse gastrointis tim necrosis, caused by cardiac arrhythmia. this is a case of a 64-year-old male, admitted to the hospital

with a history of progressive cough, undergone right lobectomy for hepatocellular carcinoma. after 14 months multiple metastases of left adrenal gland. serum protein induced by vitamin k absence, patient treated with 5 with carboplatin, two more

cycles of chemotherapy, third line chemotherapy with oral capecitabine. patient was transferred in february of 2009. this is blood work. chest radiograph revealed milliary nodules, ct scan indicated abdominal lymph nodes.

you can see the chest x-ray with the long ct scan around the lymph nodes. palliative care, seen every three months, continued to have the cough and dyspnea, no change with chest radiograph. in september of 2009 the patient had improved and the cough and

dyspnea stopped. chest radiograph revealed all metastatic nodules disappeared. this is the x-ray. serum afp dropped, whole body pet scan, lesions had disappeared. the patient said he did differently, he had taken a

herbal medicine for approximately one week on recommendation of his family, all the biomarkers were down. you can see february of february, june, september, nodules are clear. chest was clear. all the blood work was within

normal limits. so was this spontaneous regression or what he took? i don't know. this 12-year-old girl was referred to pediatric department, she had diarrhea and weight loss. she had family history of second

degree relative with colorectal cancer, iron deficiency, anemia, low albumin, she had a colonoscopy, bleeding with poly poid mass occupying greater than 50% of the lumen, squamous cell carcinoma. this is just the tumor here. this is ae-1 and ae-3 and some

of the other markers. she was treated, ct scan showed reduction in tumor volume with thickening of the parietal rectum wall and decrease in size of lymph nodes. there was an absence of intraluminal mass, negative for malignancy.

spaining for p63 was negative. six months after pet/ct revealed progression without local progression. she had a colonoscopy that did not show changes in the rectal mucosa, biopsy results negative. following that metastasis without squamous cells.

she was diagnosed adenosquamous carcinoma of the rectum, rapidly progressed with liver metastasis, enlargement of gastrohepatic lymph nodes, treated with chemotherapy, negative for 12, 13 and 61 codons of the k-ras genes, treated with bevacizumab and

three months of treatment patient was stable metastatic disease without toxicity reported. so that's the tumor there, with the initial treatment there was some improvement, metastasis in the liver and lymph nodes. so are case reports important in

the era of evidence-based medicine? some people believe that case reports are not important because a single case report cannot provide evidence of efficacy or safety in diagnosis and treatment of a disorder. others look at a hundred case

reports and the in the archives of determine nothing and there was support of public results exaggeration of claims of efficacy and safety, inadequate informed consent reporting and underreporting of patient-centered outcomes. from a business perspective case

reports are rarely cited. they negatively impact the -- negatively affect the impact factor, and quite a number of researchers -- this may have an impact on companies, which will use their dollars to advertise their products. so evidence-based medicine is

finding the best evidence for clinical decisions, for example what particular therapy or diagnostic test do we apply to a particular patient? whereas randomize trial gives an evaluation of therapies or tests especially when the clinical value is not clear cut.

however case reports are equally important in the progress of medical science and education. review, an example is nayfield and gorin that looked at tamoxifen-related ocular toxicity. as a result they were able to get a clear picture of the

nature and distribution of the toxicity, as well as severity of the ocular findings. they recognized difficulty in attributing ocular findings to tamoxifen and other competing causes of retinal, macular and corneal abnormalities.

nordin conducted review, leading to more appropriate outline of patient characteristics, able to hypothesize on the underdiagnosing of cases, causes of treatment failure, and lack of controlled trials but also they found usefulness of the second look laparotomy.

we looked at case reports of they were used by cancer patients, using case reports had apparent antitumor effect, 21 reports of toxic effect. clinical trials in green tea, phyto estrogens, mistletoe, and studies done for saw palmetto, many have either not yet been

explored or results have not been reported in english because we confine review to english in meta-analysis of case reports or case series only characteristics and outcomes can be better assessed at this level of clinical research. in situations where there's no

adequate evidence the best of the first line evidence must be taken into account. impact of case report on medical research, taking away from this table is that case reports followed by clinical trials 17% of case reports followed by clinical trials, this looked at

case series and the same question, case report followed published trials, 31%. how do you write a good case report? some smart people came together, the care group, they gave guidelines, this suggested that

a good case report should have a title, key words, abstract, introduction, patient information, clinical findings, time line of treatment, diagnostic assessment, therapeutic intervention, what were the follow-up and outcomes, have a good discusion, include

the patient perspective, and the informed consent. so what does it take away from this case report lecture? if case reports will be useful they should be written in a structured manner which the care group recommended. but case reports are relevant in

medicine today, case reports led to the clinical trials that they are doing now. in rare instances where clinical trials are not possible for ethical reasons case reports or tee residual may be the only evidence available to recommend treatment.

thank you for coming and listening, and i thank my colleagues, dr. white and dr. zia. any questions? yes, ma'am. >> well, from their view, they believe that when we write case report, the author, the way to

see the patient's perspective so that i guess that would go along the area of patient's reported outcomes, this is what the patient felt about the treatment that they have gone through. they wanted to include that in case reports, like i said, it's just a recommendation guideline,

nobody -- i have yet to see a case report that has -- >> my second question is -- gnawed (inaudible) >> generally the same level to generate new hypothesis, information that makes you go for something else.

if the generated information that will generate hypothesis, yeah. any other questions? thank you very much.