Big data yield new hopes, chal­lenges in can­cer care

Modern Healthcare - - NEWS - By Sabriya Rice

The abil­ity to se­quence the genome of thou­sands of can­cer­ous tu­mors and com­pile the data in huge ware­houses is giv­ing sci­en­tists un­prece­dented tools for un­der­stand­ing the ge­netic roots of can­cer. That ca­pa­bil­ity al­lows them to iden­tify who may be pre­dis­posed to de­velop the disease. It also gives drug com­pa­nies and clin­i­cians new tar­gets for fight­ing, man­ag­ing, and in a hand­ful of cases, cur­ing some of the es­ti­mated 200 forms of can­cer.

In­deed, the ex­plo­sion of can­cer ge­nomic data prompted the White House to launch its Na­tional Can­cer Moon­shot, a bil­lion-dol­lar ef­fort led by Vice Pres­i­dent Joe Bi­den to “elim­i­nate can­cer as we know it” and “make a decade worth of ad­vances in five years.” But while sci­en­tists are en­thu­si­as­tic about their in­creased ca­pa­bil­i­ties to make progress in the fight against what one au­thor has called “the em­peror of all mal­adies”—can­cer now ri­vals heart disease as Amer­ica’s No. 1 killer with 589,000 deaths a year—the breadth and depth of the data be­ing gen­er­ated are pos­ing huge chal­lenges for re­searchers, on­col­o­gists and pa­tients.

The data are com­pli­cated to un­der­stand, and the process for data col­lec­tion is un­struc­tured, de­cen­tral­ized, and some ar­gue, only gath­ered in si­los. And much of what is col­lected re­mains un­an­a­lyzed be­cause there is lit­tle con­sen­sus on the best ways to pur­sue data-driven re­search. In­deed, the cur­rent wave of op­ti­mism about new dis­cov­er­ies that could lead to

Can­cer now ri­vals heart disease as Amer­ica’s No. 1 killer, but the breadth and depth of the data be­ing gen­er­ated are pos­ing huge chal­lenges for re­searchers, on­col­o­gists and pa­tients.

cures might ul­ti­mately be pulled out to sea by an un­der­cur­rent of unan­swered ques­tions.

The en­vi­ron­ment “is ex­cit­ing, but frus­trat­ing,” says Dr. Bar­rett Rollins, chief sci­en­tific of­fi­cer at the Dana-Farber Can­cer In­sti­tute in Bos­ton. “While we all see what big data could po­ten­tially do, we’re frus­trated be­cause we don’t yet have all the tools or fully un­der­stand how to use it.”

It’s a time of “pro­found tur­bu­lence,” adds Dr. Julie Vose, pres­i­dent of the Amer­i­can So­ci­ety of Clin­i­cal On­col­ogy, in her in­tro­duc­tion to the group’s re­cent re­port, The State of Can­cer Care in Amer­ica.

The re­port un­der­scores the strug­gle to make sense of the “in­for­ma­tion over­load” of ge­nomic data. It states that clin­i­cians don’t have an­swers to the many ques­tions be­ing raised about the risks and ben­e­fits of ge­netic screen­ings and the role they play in se­lect­ing treat­ments.

In 2015, the Food and Drug Ad­min­is­tra­tion ap­proved molec­u­lar di­ag­nos­tic tests for lung and col­orec­tal can­cers, a tremen­dous ben­e­fit for some pa­tients with those di­ag­noses. But while such tests can change the course of treat­ment for some vari­ants of those dis­eases, for the ma­jor­ity of can­cers, the ge­nomic tests gen­er­ate in­con­clu­sive re­sults or re­veal ab­nor­mal­i­ties for which clin­i­cal guide­lines and treat­ments do not ex­ist.

Mul­ti­dis­ci­plinary teams of clin­i­cians, sur­geons, ra­di­a­tion on­col­o­gists, ge­neti­cists and oth­ers have formed molec­u­lar tu­mor boards at some re­search in­sti­tu­tions. They

study these ab­nor­mal­i­ties and of­fer guid­ance to clin­i­cians about which tu­mor types are worth test­ing and at which stage of the disease. But the op­por­tu­ni­ties to use ge­nomic test­ing to in­form clin­i­cal prac­tice re­main few and far be­tween. From the first se­quenc­ing of the first hu­man genome in 2003—con­sid­ered by Mod­ern Health­care read­ers as the most sig­nif­i­cant med­i­cal break­through of the past 40 years—can­cer pa­tients, their on­col­o­gists and the re­search com­mu­nity have be­lieved the new tech­nol­ogy would pave the way for un­der­stand­ing the ge­nomic un­der­pin­nings of dozens of can­cers.

“The op­por­tu­nity space sud­denly ex­ploded,” says Adam Res­nick, found­ing co-di­rec­tor of the Chil­dren’s Hos­pi­tal of Penn­syl­va­nia’s new Cen­ter for Data-Driven Dis­cov­ery in Bio­med­i­cine. Se­quenc­ing the genome “pro­duced a trans­for­ma­tion in the sci­en­tific arena,” he says.

In­deed, the ad­vanced un­der­stand­ing of ge­nomics led the drug in­dus­try to de­velop some new treat­ment op­tions. There are now more than 100 tar­geted can­cer drugs, which work by block­ing the ac­tion of spe­cific genes and their ex­pressed pro­teins in can­cer­ous cells, as well as im­munother­apy agents, which mo­bi­lize the body’s im­mune sys­tem to kill can­cer cells with spe­cific ge­netic mu­ta­tions. The main­stays of can­cer treat­ment for decades—chemo­ther­apy, ra­di­a­tion and surgery—are now be­ing re­placed in some cases by more pre­cise ap­proaches. Of the more than 60 can­cer drugs that the FDA ap­proved be­tween 2005 and 2014, nearly 67% were pre­ci­sion medicine ther­a­pies, ac­cord­ing to the Amer­i­can So­ci­ety of Clin­i­cal On­col­ogy, or ASCO. And 12 of the 15 can­cer drugs ap­proved in 2014 were tar­geted ther­a­pies.

Some ma­jor can­cer cen­ters on the lead­ing edge of treat­ment and re­search now se­quence the tu­mor of ev­ery can­cer pa­tient and use that data to choose the pa­tient’s treat­ment. At the Dana-Farber Can­cer In­sti­tute, for ex­am­ple, the re­sults of DNA tests in­flu­ence clin­i­cal care for nearly 3 out of ev­ery 4 pa­tients. But just 19% of those pa­tients re­ceive an FDA-ap­proved drug. For another 54%, the se­quenc­ing only pro­vides in­sight into whether they might qual­ify for a clin­i­cal trial, Rollins says.

The quest now is to ex­pand the par­a­digm, so that more than just a mi­nor­ity of can­cer pa­tients can get en­rolled in tri­als. “The only peo­ple who get this de­tailed se­quenc­ing are ei­ther rich enough to travel to a ma­jor can­cer cen­ter or just hap­pen to live near one,” Rollins said. “In­sur­ers should strongly con­sider re­im­burs­ing for it ... which will de­moc­ra­tize the process.”

Some in­sur­ers are al­ready mov­ing in that di­rec­tion. In Jan­uary, In­de­pen­dence Blue Cross, a Blue Cross and Blue Shield af­fil­i­ate based in Philadel­phia, said it struck a deal with health tech­nol­ogy firm Nan­tHealth to of­fer cov­er­age for genome se­quenc­ing. Com­mer­cial plan mem­bers with can­cers that did not re­spond to tra­di­tional ther­a­pies gained ac­cess to the test­ing in March.

The steady growth of new can­cer ther­a­pies based on ge­nomics has added to an al­ready huge prob­lem for prac­tic­ing on­col­o­gists. De­ter­min­ing the best course of treat­ment for a can­cer pa­tient has be­come a daunt­ing task. A re­cent pol­icy state­ment from ASCO called the tens of thou­sands of ex­ist­ing clin­i­cal path­ways “over­whelm­ing and un­sus­tain­able.”

The si­los within which many re­searchers work on new can­cer ge­nomics pose another prob­lem. As can­cer cen­ters and their molec­u­lar re­view boards eval­u­ate pa­tients, they need to form col­lab­o­ra­tions in or­der to gen­er­ate a crit­i­cal mass of data for mean­ing­ful anal­y­sis. “To un­der­stand what ex­actly is tak­ing place, we have to po­si­tion that pa­tient in the com­par­a­tive space of pa­tients with sim­i­lar and dif­fer­ent forms of can­cer,” Res­nick says. “That’s the only way you can draw in­ter­pre­ta­tions.”

Can­cer re­searchers are be­gin­ning to co­or­di­nate across in­sti­tu­tions. They’re also seek­ing ways to quickly dis­sem­i­nate their find­ings to on­col­o­gists in com­mu­ni­ties with­out a can­cer cen­ter. Com­mu­nity on­col­o­gists treat about twothirds of can­cer pa­tients. But they of­ten lack ac­cess to the lat­est re­search and re­lated ad­vances, ac­cord­ing to a re­cent com­men­tary in the New Eng­land Jour­nal of Medicine by Dr. Fran­cis Collins, di­rec­tor of the Na­tional In­sti­tutes of Health, and Dr. Dou­glas Lowy, act­ing di­rec­tor of the Na­tional Can­cer In­sti­tute.

ASCO would like to en­list com­mu­nity on­col­o­gists in gen­er­at­ing another big-data stream that could be used for im­prov­ing can­cer care—in this case, get­ting the most out of ex­ist­ing ther­a­pies. Last year ASCO launched CancerLinQ, a project that cap­tures data about ge­nomic tu­mor types, treat­ments and out­comes. The idea is to an­a­lyze pat­terns in that data to de­ter­mine which treat­ment pro­to­cols have worked best for which pa­tients. That in­for­ma­tion could en­able more per­son­al­ized treat­ment for fu­ture pa­tients.

A moon­shot can only be at­tained by amass­ing data and an­a­lyz­ing all can­cer pa­tients who come into physi­cians’ of­fices, “not just that small mi­nor­ity who are en­rolled in ran­dom­ized con­trolled tri­als,” says Kevin Fitz­patrick, CEO of the CancerLinQ project.

CancerLinQ of­fi­cially rolled out in June 2015 with 15 par­tic­i­pat­ing on­col­ogy prac­tices. While ASCO re­ceived more than 190 in­quiries from groups want­ing to par­tic­i­pate, a spokesper­son told Mod­ern Health­care that so far only 40 have signed con­tracts to join. They cite pa­tient-pri­vacy con­cerns and dif­fi­cul­ties with the tech­ni­cal in­ter­face. They also worry about the work­load re­quired to in­put all the data.

Dana-Farber sup­ports the ef­fort but says its can­cer cen­ter is not par­tic­i­pat­ing be­cause for now “the ben­e­fits of join­ing are not yet worth the per­son­nel costs,” Rollins said. He is on the steer­ing com­mit­tee of a sep­a­rate project run by the Amer­i­can As­so­ci­a­tion for Can­cer Re­search, called Project GE­NIE. Seven can­cer cen­ters are par­tic­i­pat­ing in that project, which aims to mine data on tu­mors that have been se­quenced, in or­der to iden­tify trends and pro­vide re­searchers with more of the ba­sic sci­ence that ex­plains how can­cer hap­pens.

Im­munother­apy is another area in which big data is ad­vanc­ing re­search. In March, Johns Hop­kins Medicine launched a new can­cer re­search cen­ter to fo­cus specif­i­cally on im­munother­apy. It re­ceived $125 mil­lion in fund­ing, in­clud­ing $50 mil­lion given by for­mer New York City Mayor Michael Bloomberg and another $50 mil­lion from phi­lan­thropist and movie pro­ducer Sid­ney Kim­mel.

Dr. Drew Par­doll, the in­au­gu­ral di­rec­tor of the new Bloomberg-Kim­mel In­sti­tute for Can­cer Im­munother­apy, hopes the cen­ter will gen­er­ate in­no­va­tive clin­i­cal tri­als and global col­lab­o­ra­tion with sci­en­tists and can­cer cen­ters. Fo­cused re­search “will ad­vance im­munother­a­pies to the point where the im­mune sys­tem will ul­ti­mately be able to beat 100% of can­cers,” Par­doll said. “The po­ten­tial to con­trol or cure even the most ad­vanced treat­ment-re­sis­tant can­cers has been elu­sive un­til now.”

The big­gest chal­lenges come down to amass­ing the in­cred­i­ble amount of knowl­edge gained about can­cer and trans­lat­ing that knowl­edge into new ther­a­pies that can help pa­tients and pro­vide new guide­lines for the on­col­o­gists who treat them.

But some ex­perts worry that the rush to re­search the pos­si­bil­i­ties gen­er­ated by big data places too much em­pha­sis on de­vel­op­ing break­through drugs and other ther­a­pies on the “ex­otic ends of the spec­trum,” says Dr. Leonard Licht­en­feld, deputy chief med­i­cal of­fi­cer of the Amer­i­can Can­cer So­ci­ety. Such ini­tia­tives should not di­vert at­ten­tion from a fo­cus on pre­ven­tion, he says. That con­cern was shared by the deans of more than 70 pub­lic health pro­grams, who sent a let­ter to Vice Pres­i­dent Bi­den in March. Hype over the can­cer “moon­shot” may be un­der­valu­ing the roles that pub­lic health and pre­ven­tion play. “His­tory has shown that the great­est im­pact in re­duc­ing can­cer mor­tal­ity rates has come from pre­vent­ing can­cers,” the let­ter stated.

Cu­ra­tive treat­ments of­ten ap­pear more ex­cit­ing to the pub­lic, and they are es­sen­tial, but con­trol­ling can­cer is also a pol­icy and pub­lic health chal­lenge. “We must op­er­ate on both fronts,” the let­ter said.


Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.