Cancer Renal

[ silence ] >> i'm really very honored to be part of thisgreat symposium as a tribute to joe. in 1982 shortly after i started my career i purchasedthe first edition of schottenfeld and fraumeni-- the textbook "cancer epidemiology and prevention."and it became my inspirational bible. i learned everything i knew about cancer epidemiologyfrom that book. little did i dream that i would actually get to meet the editors ofthat book, that i would actually be invited to write a chapter in the third edition, thati would get to work at dceg as an ipa with joe, and that i feel privileged to be ableto call joe not only a role model and mentor but a colleague and a friend. we in the extramuralcancer epidemiology community who were struggling

to build cancer epidemiology programs in cancercenters looked with awe to what joe created at the nci. he was able to assemble and inspirea powerhouse of cancer epidemiologists who were and still are the leaders in the field,both nationally and internationally, and every time i go to dceg i feel that i'm in the presenceof giants and not just because of my own short physical stature. so, joe, on behalf of theentire extramural cancer epidemiology community, we thank you very much for making us try harder.and now i think we're moving from pedigrees to populations, and i'm going to be introducingthe two speakers. and someone with a very wicked sense of humor selected me to introducethe first speaker. in the interest of full disclosure, i must state that dr. engels,besides his many other achievements, is the

father of two magnificent children who justhappen to be my grandchildren. hubert humphrey said something which actually stephen saida slightly different version. he said, "behind every successful man stands a proud wife anda very surprised mother-in-law." well, i am not a surprised mother-in-law. eric has donea fantastic job of describing cancer epidemiology in immunosuppressed individuals, both thosewith hiv/aids and those in transplant recipients, and he has done this by doing superb linkagesbetween different registries which is a model for perhaps future descriptive epidemiologicresearch as well. so, i have great pleasure in calling on eric to talk about transplant-associatedimmunosuppression and cancer. [ applause ]

>> well, thank you very much, margaret, forthat kind introduction. and it is very intimidating being introduced by one's mother-in-law, butyou didn't say anything too embarrassing. [ inaudible comment ] and it is indeed a great pleasure to be speakinghere today in honor of joe fraumeni and his many accomplishments in the scientific worldand for his establishment of a very productive scientific community in dceg. so, today i'llbe talking about a number of topics in organ transplantation and cancer. i'll start witha history of organ transplantation to put joe's contributions into context, and alsoby way of background to describe our transplant cancer match study. i'll then talk about thespectrum of cancer risk in transplant recipients

and focus on non-hodgkin lymphoma which isone of the most common cancers, and then i'll close with some observations about the effectof immunosuppressed populations on the overall cancer burden in the united states. so, thedream of transplanting-- taking a diseased organ out of a person-- and putting a healthyorgan in has been with medicine probably since its beginnings. perhaps the first organ transplantwas performed by two brothers-- cosmas and damian-- who were early physicians and religiousmartyrs, and they are reported to have transplanted the leg of a deceased donor onto a man sufferingfrom gangrene in the year 300, but there are some anomalies with this case report. first,the donor had been dead and buried by four days at the time that this transplant wassupposedly done. second, the transplant surgeons

took the diseased leg from the recipient andtransplanted it back onto the deceased donor, and that's generally not what's done today.the third anomaly is that, as this picture hints at, these surgeons were themselves deceasedalready at the time of the operation, and finally, perhaps more important for our talktoday, the recipient was not given any immunosuppressive drugs, and as i'll show you, that's a keyfeature necessary for successful transplantation. so, that story is a bit apocryphal, but thereal pioneers of organ transplantation are shown here. beginning in the late 1800s andearly 1900s there were attempts at transplantation first in animals and then in some human cases,but these were all complicated by rejection of the donor organ. in 1954 joseph murrayconducted the first successful kidney transplant

between identical twins. the 1960s markedthe development of successful immunosuppressive therapy by thomas starzl, mainly based onazathioprine, and then in the 1960s and 1970s starzl pioneered successful liver transplantationin humans. and then in 1967 christiaan barnard performed his first successful heart transplant.now, this slide shows the number of transplants performed in the united states in each calendaryear, and it shows a progression from when transplantation was experimental therapy towhere it's now become the standard of care for people with endstage organ disease. twoother landmarks are shown here-- one is the introduction of cyclosporine which was a majoradvance in immunosuppression, and then the initiation of the us transplant network in1987 which allowed for regional sharing of

donor organs and facilitated the establishmentof a national program. now, i want to put these landmarks into context and show youanother landmark, and that is when the first epidemiologic study of cancer in transplantrecipients was performed by these two gentlemen-- bob hoover and joe fraumeni. this was in 1973--40 years ago-- and well before the time when transplantation was a major part of clinicalmedicine. now, why did doctors hoover and fraumeni show an interest in this topic? well,that really gets back to the issue about rejection-- the major barrier to transplantation. thisoccurs because a recipient's immune system recognizes the donor organ as foreign, andthis is mediated by t-cells and it leads to a gradual destruction and loss of utilityof the donor organ. and so, transplantation

requires the use-- the lifelong use-- of immunosuppressiontherapy which blocks t-cells from attacking the donor organ. unfortunately, there is anadverse side to this process, and that is that immunosuppression leads to some complications.it leads to serious opportunistic infections, and some of these are very similar to thosethat people with aids get. in fact, transplant recipients' immunosuppression mirrors in manyways what we see in people with aids. there was also an important hypothesis that waspromulgated beginning in the 1950s that an intact immune system protects us from developingcancer. this was proposed by burnet, and it was known as the immunosurveillance hypothesis.and so, one would therefore hypothesize that immunosuppressed people would have an excessof cancer, and indeed there were case reports.

as transplantation came to be a clinical realitythere were case reports of unusual cancers occurring in transplant recipients, and itwas these reports that motivated bob and joe to study cancer in transplant recipients.so, here is their paper published in the lancet in 1973. this was the first study not onlyof cancer in transplant recipients but the first epidemiologic study of cancer in anyimmunosuppressed population. they followed up more than 6,000 transplant recipients atover 200 institutions throughout north america and europe, and the key findings are shownin this table. there was an excess risk of one particular type of cancer, and that wasnon-hodgkin lymphoma and one subtype of nhl in particular-- reticulum-cell sarcoma. so,here you see that the relative risks are 280

for males and 700 for females compared topeople in the general population. these are really profoundly increased relative risksand point to the importance of immunosuppression in causing these cancers. now, reticulum-cellsarcoma is not a terminology that we use anymore. these are large cell lymphomas, and most ofthese would have been diffuse large b-cell lymphomas. and those cancers are mostly causedby epstein-barr virus as we now know, and lack of immune function leads to replicationof ebv infected cells and the eventual development of cancer. there was not a marked excess ofother cancers, and this was a partial refutation then of burnet's most general immunosurveillancehypothesis. there was a specific excess for lymphoma. now, following that landmark studythere were other studies of cancer in transplant

recipients as this became more of a part ofclinical practice, and those studies are shown in this table. and by and large they certainlyshowed an increased risk for non-hodgkin lymphoma and probably other virus-related cancers,and they all showed an increased risk of skin cancers, especially squamous cell skin cancers.but it was uncertain if the risk was elevated for other cancers, and that's because eventhough these studies are large-- you can see the number of transplant recipients here--they really weren't large enough to provide robust estimates of risk for most cancersthat we would be interested in. another limitation was that most of these studies only includedkidney recipients, but by this time successful transplantation of other organs was part ofclinical practice. so, that was the status

of the field in 2005, and it was at that timelooking over this body of work that we decided to launch the transplant cancer match study.and this is a linkage of the us solid organ transplant registry with a number of cancerregistries in the united states, and this study has some strengths that go beyond theprevious work. it is population-based. it includes recipients of all solid organs, notjust kidneys but including people who got livers, hearts, or lungs. it has data on recipientrisk factors that are recorded in the transplant registry, and we had complete ascertainmentof cancers as recorded in the cancer registries. moreover we linked not only the recipientsbut we linked the wait list candidates and donors so we had information on cancer inthose populations. here at nci this study

has been led by me along with ruth pfeiffer,my colleague in the biostatistics branch, and with joe, and this has also been a closecollaboration with hrsa-- the health resources and services administration-- which is oursister agency that oversees organ transplantation in the us as well as the transplant and cancerregistries. now, here is a summary of some of the features of our transplant populationunder study. to date we've linked the transplant registry with 15 cancer registries shown inpink. that includes data on 231,000 transplant recipients or 47 percent of the us total.so, the study is larger than the combined populations of all previous studies of transplantrecipients, and as you can see by the bar chart on the right, we have data not onlyon kidney recipients-- 130,000 of them-- but

50,000 liver recipients and about 35,000 thoracicorgan recipients. now, we've used these data for a few studies so far. our first paperlooked at the spectrum of cancer risk in transplant recipients, and i'm going to show you themain results from that study now. so, this table presents risks for virus-related cancers,and it shows the relative risk for each cancer compared to the general population. the mostcommon cancer was non-hodgkin lymphoma as expected, and the relative risk was sevenand a half times higher than seen in the general population and the risk was also increasedfor hodgkin lymphoma with a relative risk of 3.6. both of those types of cancers arecaused largely by epstein-barr virus. there was also a markedly increased risk for kaposisarcoma. this is one of the hallmarks of aids,

and it's also increased to a great degreein transplant recipients. here the relative risk is more than 60, and the risk was increasedfor the hpv-related cancers-- the human papillomavirus-related cancers. here you see relative risks in therange of two to seven with the notable exception of cervical cancer where the risk is not increasedat all, and we suspect that that is due to effective screening of women who have an organtransplant. the risk is also increased for other cancers not related to viruses. hereyou see some common cancers-- lung cancer, kidney, melanoma, and thyroid cancers-- therelative risks are all in the range of two to four. lip cancer is very much increased.this is probably because it's very similar to skin cancer, and here the relative riskis 16. the risk is increased for colorectal

cancer but not to the same degree-- a relativerisk of only 1.2-- and the risk is not increased at all for some other common cancers-- prostate,breast, and ovary. so, you can see here that there's a wide spectrum of cancer that transplantrecipients are at risk for, but it's not all cancers. now, i want to turn to talk a littlebit about non-hodgkin lymphoma, and i'll describe some results related to the risk in kidneyrecipients related to function of their transplanted organ, the risk for individual subtypes ofnhl, and risk at extranodal sites. so, one of the unique features of kidney transplantationis that people with severe kidney disease can live for an extended period of time withouta functioning kidney transplant because of the availability of dialysis. so, what i'mshowing you here is a progression across time

for different clinical states that somebodywith kidney disease can be in, and i'm showing you the relative immunosuppression associatedwith those states. so, while somebody is on the wait list on dialysis, they might be somewhatimmunosuppressed, but when they get their transplant and they go on immunosuppressivemedications that immunosuppression becomes much more profound. some people lose theirkidney and go back on dialysis so the immunosuppression decreases, and then some are fortunate enoughto get another transplant so it goes up again. and then that transplant might fail so itmight go down. so, this provides a model if you will-- an actual experiment-- for lookingat the effects of immunosuppression, especially for cancers that might have a short latency--a very tight association-- with immunosuppression.

and we've been doing these analyses now fora number of cancers, and here are the results for nhl. and you can see a very marked sawtoothpattern where the risk increases and then decreases in concert with the functioningof a transplanted kidney. so, this is compelling evidence that immunosuppression here is drivingthe development of these cancers. we also know that non-hodgkin lymphoma is not onecancer. it's a range of cancers with different clinical and pathologic features, and so we'velooked at the risk of nhl for different subtypes. diffuse large b-cell lymphoma is the mostcommon type, and the relative risk compared to the general population is quite high at13. the risk is also elevated for burkitt lymphoma. both of those types are relatedto epstein-barr virus. the risk is increased

for marginal zone, lymphoplasmacytic, andt-cell lymphomas, and the elevation in these cancers really points to the importance ofeither viral infections or immune disturbance in their etiology. in contrast, the risk fortwo common subtypes in the general population-- follicular lymphoma and chronic lymphocyticleukemia-- you can see that these are not increased at all in transplant recipientsso it points to a different etiology. and here are the results that i mentioned aboutthe location-- the site of origin-- of the diffuse large b-cell lymphomas. here i'm showingyou relative risks for diffuse large b-cell lymphoma, separately for kidney recipients,liver recipients, heart and lung recipients, and we first focus on nodal lymphomas-- diffuselarge b-cell lymphomas that arise in the lymph

nodes. and the relative risks here are quiteelevated-- around tenfold-- compared to the general population, but if you look at lymphomasthat arise in the transplanted organ, here the relative risks are really quite high--more than 100-- for kidney, liver, and lung recipients. there are no cases of heart lymphomasbecause these are just so rare, and then if you look at the diffuse large b-cell lymphomasthat arise at other extranodal sites, these risks are increased but not to the same degree.so, this pattern-- this very high risk for lymphomas arising in the transplanted organs--points to the importance of immune dysfunction-- a localized form of immune dysfunction-- that'scontributing to a development of these cancers, and that suggests that rejection is involvedin stimulating the development of these cancers.

and then i'd like to talk briefly about thepopulation level impact of cancers in immunosuppressed people. so, here are the same numbers i showedyou before-- the number of transplants conducted in the united states in each calendar year--but now i'm interleaving on top of that the number of aids cases diagnosed in each calendaryear. so, the aids epidemic hit the united states in the early 1980s. the number of casesthat were diagnosed surpassed the number of transplants very quickly in the late 1980sand early 1990s. the number of aids cases has dropped over time because of the availabilityof hiv therapy that's effective in treating these patients, but what i want to point outto you is that even though the total number of aids cases in the united states is reallynot that large, these people have had a very

noticeable effect on the overall nationalcancer burden. and so, here are the number of diffuse large b-cell lymphomas in the unitedstates. these are all cases in the general population, and these have increased overtime because the rate of lymphoma has increased but also because the size of the us populationhas increased. and through a linkage study-- a separate linkage study that we have calledthe hiv/aids cancer match study-- we've identified which cases of these are in people with hiv,and 6 percent of all of these cases-- the yellow cases here-- since 1980 have been inpeople with hiv. so, this is an impact that is out of proportion to the size of the hivpopulation. if you look at kaposi sarcoma, the numbers are much more striking. here becauseof the very high risk of ks in people with

hiv you see that more than 80 percent of allks cases in the united states since 1980 have been in people with hiv. now, we speculatethat the impact of the transplant population on cancer burden will also be out of proportionto the size of this population, but we haven't yet done those analyses. so, what insightshave we learned? well, immunosuppression increases the risk for some cancers, especially virus-relatedcancers, but not all cancers. the patterns of risk differ for different subtypes suggestingthat the etiology varies across subtypes and that the cancer risk in a small populationcan have major effects at the general population level. our next steps are to try to identifyrisk factors for cancer in transplant recipients using the data we have from the transplantregistry. we can look at medical conditions

that these transplant recipients have andthe use of specific immunosuppressive medications. we'd also like to better understand the riskfactors for the cancers that are not related to infections and to study the differencesin the spectrum of cancer between transplant recipients in people with aids. these spectraare similar but they're not identical, and that can provide some clues. and then we'dlike to look at the impact of transplant associated cancers on the general population burden.so, with that i'd like to close by thanking my collaborators on the transplant cancermatch study. this has been a tremendous undertaking. i'd especially like to single out today joefraumeni's contributions. he has been instrumental in the conception, design, and implementationof this study. his encouragement, his belief

that this work is important, has helped motivateus over the long course of this work, and i'd especially like to thank him personallyfor the mentorship that he provided me in my career so far. so, thank you, joe, andcongratulations.