OUTLOOK
`t-Iuemr'j‘hiiiu
`
`IHE INTERNATIONAL WEEKLY JOURNAL OF SCIENCE
`
`Anti tumour immunity
`enhanced by inhibiting
`PD—Ll XPD — I and identifying
`mutant new antigens
`
`natu .
`
`
`
`_-_-"Ill'_'
`
`vi!
`
`'
`
`I
`
`[Jr's
`
`.
`
`I
`
`I
`
`_
`
`v”.
`
`_-III._.
`
`IMMUNE-CHECKPUINT
`
`BLUBKADEIN
`
`, 7 DO NOT REMOVE FROM
`CURRENT PERIODICALS
`
`"
`
`ROOM
`
`PEER REVIEW
`
`MICROSCOPY
`
`ENERGY
`
`9 “m"HE-cm‘IiWUHE
`23" Novembet 20M
`
`ACCEPT YOUR
`OWN PAPER
`Howsume SL‘iEIlEiSI‘SEIi’e
`duping the system
`mm"
`
`THE CASE FOR
`AIMING HIGHER
`Atomic resolution is there
`for the taking
`"“4“?
`
`SIDUUUS $12.99CM-I
`‘NIGHT—TIME,
`4 8>
`.
`.
`COOI INC RV DAV
`GenOme & CO. V. UI'IIV. Of Chlcago
`i‘udiut PG R201 9-00002
`UN IV. CHICAGO EX. 2049
`
`

`

`m—REVIE.WS MICROBIOLOGY
`
`IMatching
`genomic
`sequence
`
`../'
`
`Genomic DNA
`
`RYAY
`
`
`
`
`
`
`DEC 0 3 2014
`
`ngNA
`
`umwsnsmr LIBRARY
`
`FREE POSTER
`
`CRISPR—Cas: extraordinary editing
`
`The CRISPR—Cas system evolved to provide adaptive immunity in bacteria and archaea against
`
`invading genetic elements. such as viruses and plasmids. The mechanism of interference is based on
`
`CRISPR RNAs. which are derived from foreign DNA and mediate the detection of invading nucleic acids.
`
`followed by degradation by the Cas proteins. This poster presents the molecular mechanism of CRISPR—
`
`Cas interference and highlights the use of specifically engineered guide RNAs and the Cas9 nuclease for
`
`the editing of virtually any desired genome, in addition to other applications such as imaging in live cells
`
`and the regulation of gene expression.
`
`
`
`0 Download the poster at:
`www.natu re.com/n rmicrofposters!crispr
`
`flitlllflf
`Reviews
`
`
`
`
` ;_'.'_-'._;;-
`
`-_-_-.
`
`,J (JRII‘JENI
`
`Produced with support from
`
`g 0RIGEN E
`
`Your Gene Company
`
`I
`
`nature publishing group @
`
`

`

`CONTENTS
`
`2'? November 2014 J Vol 515 l Issue No 7528
`
`THIS WEEK
`
`NEWS IN FOCUS
`
`COMMENT
`
`48? MICROSCOPY
`Hasten high
`resolution
`Stephen J
`Pennycookde .
`ge’nglVKa’mm
`recrsion
`microscopes are
`needed to analyse
`atoms effectively
`
`
`
`Bflflls 8. INS
`
`
`
`
`-
`
`‘
`
`'
`
`..
`
`_
`r’ J
`.3!
`'
`‘7 A
`
`
`
`-
`.
`473 EMISSIONS
`US—China deal renews hopes fortaiks
`474 Mi'iRINE SCIENCE
`Data-managementissues delay
`observatory project
`478 NEUROSCIENCE
`FDAconsiders howto regulate
`brain-linked devices
`
`:
`
`.
`I
`'
`I
`
`Elll‘l’llflllls
`465 CLIMATE
`Agree to agree
`Other nations must follow the lead of
`the United States and China
`465 INFECTIIJUS DISEASE
`,
`Ebola opportunity
`Slowdown should be putto good use
`4E5 SPACE
`Moon on a stick
`.
`.
`Crowdfunded mission worth a try
`it
`"Dal.” VIEW
`hr
`45? Open access
`is tiring out peer
`m3“?
`a an ms
`Scholarly system must
`adapt to workload
`
`
`
`
`
`HESEIICH HIGHLIBHTS
`463 SELECTIUNS FROM THE
`SCIENTIHC UTEHMUHE
`
`Termite queen rules roost / North China
`earthquake history / Old papers back in
`vogue / Tagged fish predated / Remote
`
`SEVEN IIAYS
`4m THENEws IN BRIEF
`Oil pipeline rejected / Charity sells
`royalty rights / Holt to head MAS /
`LHC data go open access / Wave—power
`
`companyenters administration /‘Polar
`
`code‘ to regulate ships
`
`C A R E E R 5
`
`597 DWERSITY
`
`Structural approach
`Materials scientists are tryingto
`broaden the range of people
`the field attracts
`
`.
`
`'
`
`4?? M£DICINE
`Clamdewn on clinical-trial reporting
`4m PUBLISHING
`Open-data dictum notalways followed
`479 ECOLOGY
`Galapagos gift-shop closure
`shortchanges conservation
`FEITIIIIES
`
`PUBLISHINC
`
`The peer —
`.
`I'GVIGW scam
`Evidence of self-review exposes
`weaknesses in system mm
`
`4B4 NUCLEMi POWER
`Desperately seeking plutonium
`Dwindling stockpile forces US rethink
`
`
`
` 49!] URNITHCLUCY
`control of epilepsy / Greenland on ice Richard l/an Nooro‘en
`
`
`.
`_.
`'- “i." .
`'
`‘-
`‘
`Er.
`'
`MILITflHYSBIENDE
`-'
`O
`; SP 11 f
`-
`O S or
`-
`SCienCe
`Ann Finkbeiner reviewstwo books on
`
`the legacies of the cold war PIEE 48!
`
`Fowl domination
`
`Ewen Callaway
`491 CLIMflTE SCIENCE
`A climate trance
`
`CORRESPDHBEHBE
`492 Ebola virus control / Brain training in its
`infancy / China's progress in research /
`Philae's marathon achievement
`
`.
`
`.
`
`_ _
`
`Rebecca Birch
`
`FUTURES
`Bill] Ice and
`white roses
`
`2? NOVEMBER 2014 | VOL 515 | NATURE | 459
`
`

`

`CONTENTS
`
`27 November 2014 i Vol 515 l Issue No 7528
`
`RESEARCH
`
`Prime cancer
`
`
`
`
`
`
`
`:
`I
`
`
`
`ALLISONBRUCE
`
`545
`
`559
`
`temperature under direct sunlight
`A P Ramon, M A Anoma, L Zhu,
`E Rephaeii & S Fan
`
`MOLECULAR ELEOTHONlOS Design
`and fabrication of memory
`devices based on nanoscale
`polyoxometalate clusters
`C Busche et al.
`
`PALAEOBLIMATE Evolution and forcing
`mechanisms of El Nifio over
`
`the past 21,000 years
`ZLiu et al. SEE nsv P.494
`
`554 BIOPHYSICSMultiplexsingle-molecule
`interaction profilingof DNA-barcoded
`
`”News
`L Gu et al.
`555 DANCER MPDL3280A(anti-PD-L1)
`treatment leads to clinical
`
`activity in metastatic bladder cancer
`TPowles et al. SEE nsv P. 496
`
`5E3
`
`CANCER Predictive correlates of
`response to the anti-PD-Ll anti body
`MPDLSZSOA in cancer patients
`R S Herbst et al. SEE Nsv P. 496
`.
`553 DANCE“ PD: tfli‘ftf‘ge "1‘1"“?
`responses y In I
`I mg
`adaptive immune resistance
`P C Tomeh et al. sEE uov R496
`I
`I
`.
`
`.
`
`spectrometry and game sequfficing
`M Vaclav et al. 55: MN [1495
`_
`‘
`5?? CANCER Checkpoml blockade cancer
`immunotherapy targets tumour-
`SPECIlIC ("mam antigens
`M M Gubm el 8'» SEE W P- 496
`552 NEUROSCIENOE Histone H2A.Z subunit
`exchange controls consolidation of
`recent and remote memory
`iBZovKic et al.
`53; PLANT SCIENCES Epigenetic
`reprogrammingthat prevents
`transgenerational inheritance
`ofthe vernalized state
`P Crevmén et al.
`591 VIHOLOOY A structure-based mechanism
`fortRNA and retroviral RNA
`remodelling during primer annealing
`S B Miller, FZ Yiicliz. JA Lo, 3 Wang
`8. VM D’Souza
`
`2? NOVEMBER 2014'
`
`VOL 515 | NATURE | 46]
`
`NE’Al ONLINE
`493 Papers published this week at naturecom
`
`NEVIS & VIEWS
`494 CLIMATE SCIENCE
`El Nifio's variable history
`Evolution of the El Nine—Southern
`Oscillation over the past 21,000 years
`Josephine R Brown SEE LETTER P.sso
`495 MAMMALIAN EVOLUTION
`A beast ofthe scuthern wild
`A complete skull of a Madagascar:
`Eondvaanfthere
`one
`6; SEE ARTICLE P's”
`
`495 “NEE"
`Antitumourimmunity getsa boost
`Resi’onsem 'mmune‘cmkpo'm
`blockade in multiple cancers
`Jedd D Woichok it Timothy/A Chan
`SEE LETTERS PP.558—581
`
`498 ASTRONOMY
`Cosmic triangles and
`Hack-hole masses
`Estimating geometric distances to
`active galactic nuclei and quasars
`Martin Elvis SEE LETTER P528
`499 DEVELOPMENTAL BIOLOGY
`Polarize t0 elongate
`Toll receptors direct cell rearrangement
`
`”W Tape“ SEE ”mm P523
`50‘ DIET
`Food choices for health and planet
`Projected diets {inked to greenhouse.
`gas emissions and poor health
`Eiire Stehfest 55: ARTICLE ns1e
`502 IMMUNOLOGY
`Tolerance liesinthetiming
`Lck—bound co—receptor recruitment is
`rate—limiting for negative selection
`Nicholas RJ Gascoigne
`REVIEW
`_
`_
`_
`505 ECULBGY Belowgrouncl biodiversny
`and ecosystem functioning
`R D Badge” W” “a” 0"” Putts”
`IIITICLES
`512 EVOLUTION First cranial remains of a
`gondwanatherian mammal reveal
`remarkable mosaicism
`D WKiause et al. sEE Ivav R495
`
`
`
`
`.
`
`,
`:
`|
`
`I
`{Spa-m? his
`
`555549.
`.
`. --t
`
`.5
`
`.593.589.572&5?i.
`
`‘
`
`518 NUTRITION Global diets link
`environmental sustainability
`and human health
`
`523 DEVELOPMENTAL BIOLOGY A positional
`Toll receptor code directs convergent
`extension in Drosophila
`A C Par-é et al. SEE N&V R499
`
`53]
`
`LETTERS
`525 Asmononr A dust-parallax distance of
`19 megapsrsecstothe supermassive
`blacknhole m NGC 415:!
`_
`8 F Honig. D Watson, M KIShlmOm
`‘5” Him“ SEE ”8““ P493
`SPACE PHYSIOS-An impenetrable barrier
`to uitrarelatIvIstIc electrons In the
`Van Allen radiation belts
`D N Bakeret al.
`535 MATERIALS SCIENCE Metallization of
`vanadium dioxide driven by
`large phonon entropy
`J D BUda' 9t al.
`540 OPTICS AND PHOTONIOS Passive
`radiative cooling below ambient air
`
`

`

`nature
`
`27 November 2014 t V01515 1' Issue No. 7528
`
`nature@nature.com
`
`LONDON. PARIS, MUNICH,
`SHANGHAI, TOKYO. SAN FRANCISCO.
`WASHINGTON DC. NE’I'II YORK.
`BOSTON
`EoITOR-lN-CHIEE: Philip Campbell
`NATURE EDITORIAL DIRECTOR: Ritu Dhanti
`DIRECTOR. AUTHOR AND REVIEWER SERVICES: veronioue Kierrner
`EliEcurivE EDITori: Nick Campbell
`HEAD OF EDITORIAL PRODUCTION: Lucy Banhani
`CHIEF MAGAZINE EoITorI: Rosie Mestel
`CREATIVE DIRECTOR: Kelly Buckheit Krause
`THIS WEEKINEwSI’FEATURES‘. Alison Abbott. David Adam. Celeste
`Biever, Declan Butler. Ewen Callaway, Davide Castelvecchi.
`Erika Check Hayden, Matthew Crenson. Danrel Cressey.
`David Cyranoski. Elizabeth Gibney. Heidi Ledford, Corie Lok.
`Brendan Mahler. Richard Monastersky. Lauren Morelio.
`Anna Nagle. Helen Pearson. Sara Reardon. Qulrin Schiermeier.
`Jeff Tollefson. Richard \tan Noorden. M Mitchell Waldrop.
`Alexandra Witze COMMENT: Sara Abdulla. Joanne Baker,
`Rosalind Cotter. Barbara Kiser. Lucy Odling-Smee NATURE
`PDDCASTJIIIDEO: Charlotte Stoddart. Noah Baker. Kerri Smith
`NEWS AND VIEWS: Andrew Mitchinson. Jenniter Gardiner.
`Ana Lopes. Marian Turner RESEARCH: Francesca Cesari.
`Karl Zieme1i's. l-han Chou. Tanguy Chouard, Rosamund Dew.
`Alex Eccleston. Angela Eggteston. Joshua Finkelstein. Henry Gee.
`Patrick Goymer. Hanna Granroth-Wilding. Noah Gray.
`Claire Hansetl. Marie-Thérese Heemels. Magdalena Helmer.
`Andrew Jermy. Nathalie Le Bot. Barbara Made. Juliane
`Mosslnger. Leonie Mueck. Leslie Sage. Sadat Shaoan.
`Mara Silva. Magdalena Skipper. Clare Thomas, John \tanDecar.
`Liesbeth Venerna. Ursula weiss. Michael White insrsiirsr'
`REVIEWSIPERSPECTIVES: Ursula Weiss CAREERS: Karen
`Kaplan TECHNOLOGY FEATURES: Vivien Marx FUTURES: Colin
`Sullivan SUPPLEMENTS: Herb Brody. Michelle Grayson.
`Kathryn Miller STATISTICAL ADVISER: Theresa Hyslop
`SUBEDITOFIS: Colin Sullivan. Mary Abraham, Sarah Archibald,
`Nicola Bailey. Adam Born. Nicolas Fanget, isobel Flanagan.
`Paul Fletcher. Nrsha Gaind. Anne Haggait. Dinah Loon.
`Anna Novitzky. ‘r'ukie Ozawa. Katherrne Per ry, David Price,
`Jenny Rooke. David Winters. Anna ‘r‘ork PRODUCTION: Nick
`Kemp. Donald McDonald. Stewart Fraser. Susan Gray. Simon
`Gribbin. Hazel Mayhew. Ian Pope. ‘r’vonne Strong. Roberta
`Tarran, Charles Wenz ART AND DESIGN: Wesley Ferna nde 5,
`Michael Ayres. Jennifer Burns. Ffion Cleverley.
`Barbara lzdebska. Paul Jackman. Jasiek Krzysztotiak.
`Annthea Lewis. Chris Maddaloni. Chris Ryan. Nik Spencer,
`Claire Welsh JAPAN EDITION: Eiji Matsuda. Mika Ishida.
`Mitsuyo Utsugi. Sayaka Ayarne PRESS OFFICE: Alice Henchley.
`Neda Afsarmanesh, Lisa Boucher. Sarah Lahert,
`Michael Stacey. Rebecfia Walton NATURE DIGITAL: Tom Scott,
`Cianna Timbers. Alexander Thurrell. DIptI Shah, Dan Foley.
`Robert Sullivan. Matthew Carey. Luke Stayenhagen.
`Marta Rolak. Glynn Phillips, Marcin Arciuch ADMINISTRATION:
`Caroline McLean. Roseann Campbell. Emily Banham.
`Sophie Beastey. Emma Clilford. Dana Coles, Ella Kahn.
`Melissa Rose. Lauren Sequeira. Kenneth Simpson.
`Kimberly Tolteson. Michelle Viscuso. Andrea Von.
`Rebecca White. Diane Yorke
`
`'ciil'Er'ExzcuTwE'o'rrlcze: Steven lnchcoon'ibe
`PUBLISHER: Sarah Greaves
`PUBLISHER (SUPPLEMENTS 8.: PARTNERSHIPS): Richard Hughes
`PUBLISHING PROJECT MANAGER: Anastasia Panoutsou
`
` DISPLAY ADVERTISING:
`USKCANAUA: dispfay@us.nature.com
`EUROPE: display@nature.coin
`ASIA-F‘AE:-‘FIC: nature@natureasia.com
`SPONSORSHIP: sponsorship@nature.con1
`NATURE JOBS: natureiobs@nature.coni
`
`UWCANADAJLATIN AMERICA: marketing@us.nalure.com
`EU R0 PEAREST DF WON LD: ma rketing@ natu re .COm
`JAPANICHINAJ‘KO REA: marketing@n atureasiacom
`INDIA: IIIarkeiing-india@nature.com
`
`AM ERicAS: i'nstitu'tio n's@ natUrenyiconI
`ASIA-PACIFIC: institutions@natureasia.com
`AUSTRA LIAI'NEIItr ZEALANO: natu re@ma cmillan.com.au
`EU R0 Pia-“Row: institutions@ naturecom
`INDIA: npgindia@ naturecorn
`
`E.- (INCLUDING POSTAND PACKING)
`.
`...
`-
`:
`UKJEUROPEIREST OF WORLD
`subscriptions@riature.com
`FULL RATE: UK £1,942 Europe E3014
`PERSONAL RATES: UK £135: Europe €209: Rest of World £210
`STUDENT RATES: UK iii-’9: Europe €122: Rest of World £125
`Individual rates apply to payments by personal cheque or
`credit card. BACK ISSUES: £20.00.
`USAI'tCIANIIIDIMIf LATIN AM ERICA
`subscriptions@us.nature.i:om
`msnnmoNALr’ooRPORATE RATE: $3390
`PERSONAL RATE: $199 STUDENT RATE: $99: Posrooc RATE: $1 19
`Printed in USA. Individual rates available only to subscribers
`paying by personal check or credit card. Orders for
`studentfpostdoc subscriptions must be accompanied by a
`copy oI student ID. Rates apply to USA. Canada, MexicoICentral
`& Soulh America. Add 7% GST lax in Canada (Canadian GST
`number 14091 1595). BACK Issues: US$20.00.
`JAPANI'CHINAI’KOREA
`JAPAN: natureasiar’japan-subscribe
`CHINA: natureasiar’china-subscribe
`KOREA: two-Anatureasiaco mr’ko-krrsubscri'be
`
`N150 LONDON: nature@n'ature.com
`The Macmillan Building. 4 Crinan Street. London N1 QXW
`Tel: +44 (0)20 T833 £1000
`NPG NEW YORK: nature@natureny.com
`1'5 Varlck St. 9th Floor. New York. NY 10013-191?
`Tel: +1 212 F26 9200
`NPG ASIA—PACIFIC: nature@natureasia.com
`Chiyoda Building. 2—3? Ichigayatamachr. SITinjukLI-ku.
`Tokyo 162-0843
`Tel: +81 3 326? 8?51
`NPG GREATER CHINA: chlna@nature.com
`42F-10r‘1 1. The Center. 989 Changle Lu. Shanghai 200031
`Tel: +86 212422 5000
`
`A list of other NPG offices can be found at
`wwvanature.cornr’npg_fcontactr’of‘lice5.html
`
`Natrrre® {lSSN 0028-0836} published weekly by ..
`UK edition: Nature Publishing Group. a division oi Macmillan
`Publishers Ltd. The Macmillan Building. 4 Crinan Street,
`London N1 QXW. UK.
`US edition: Nature America. Inc. trading as Nature Publishing
`Group. 3’5 Varick St. 9th Fl. New‘r‘ork NY10013-191?. USA.
`Japan edition: NPG Asia-Pacific. Chiyoda Building.
`2.3}- Ichigayatamachi. Shinjuku‘ku. Tokyo 162-0843. Japan.
`Registered as a newspaper at the British Post Office.
`No part of this publication may be reproduced. stored
`in a retrieval system. or transmitted in any form
`(electronic or otherwise) without prior permission from
`permissions@natureturn. In the U5. Nature (ISSN
`0028-0836) is published weekly excepl the last week in
`December by Nature America. Inc. trading as Nature
`Publishing Group. 75 Varick St. 9th Fl. New York N‘rr
`10013-1917. USA. US Periodicals postage paid at New York
`NY, and additional mailing post offices.
`US POSTMASTER: send address changes to Nature America.
`Inc. Subscription Dept. F5 Varick St. 9th Fl. New York NY
`10013-1911“. USA.
`© 2014 Macmillan Publishers Limited. All rights reserved.
`
`464 | NATURE i
`
`\'{.I[. SIS 3 2? NOVEMBER
`
`‘Joirr
`
`naturej urnais
`
`SHARE
`
`.
`
`l Available on (In:
`
`‘ App Store
`
`..
`..
`II.IlU|L' puliiistain: group @I
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`doi:l0.10§$f_r_l_ature_1390
`
`MPDL3280A (anti—PD—Ll) treatment leads to clinical
`activity in metastatic bladder cancer
`
`Thomas P0w1es', Joseph Paul Ederz, Gregg D. Finea. Fadi S. Braiteh“, Yohann LoriotS, Cristina Cruz", Joaqnim Bellmunt7,
`Howard A. Burrisg, Daniel P. Petrylalt‘i, Siew—leng Tang), Xiaodong Shena. Zachary Boyd". Priti S. Hegdes, Daniel S. Chen3
`8: Nicholas J. \logelzang‘J
`
`pre—screen samples, 59% were from resections and 27% were from bio~
`psies. Analysis was permitted on both archived and fresh tissue. The
`time between tissue collection and starting MPDL3280A treatment is
`shown in Extended Data Fig.
`l. The prevalence of positive PD-Ll
`expression (IHC score 2 0r 3 (213)) in tumour-infiltrating immune
`cells in the pre—screened population was 27% (Fig. 1a. b). Only 4% of
`pre—screened patients had positive PD-Ll expression in tumour-infilt-
`rating immune cells and in tumour cells.
`Patients in the UBC cohort were dosed between 13 March 2013 and
`
`There have been no major advances for the treatment of metastatic
`urothelial bladder cancer (UBC) in the last 30 years. Chemotherapy
`is still the standard of care. Patient outcomes, especially for those in
`whom chemotherapy is not effective or is poorly tolerated, remain
`poor“. One hallmark of UBC is the presence of high rates of so matic
`mutations”. These alterations may enhance the ability of the host
`immune system to recognize tumour cells as foreign owing to an
`increased number ofantigens“. However, these cancers may also elude
`immune surveillance and eradication through the expression of pro—
`1 la nualy 2014. As ofthe clinical cutoffdate of 1 lanuary 2014, 68 patients
`grammed death-ligand 1 (PD-L1; also called CD274 or B7—H l} in the
`tumour microenvironment”. Therefore, we examined the anti—PD—
`with UBC received treatment and were evaluable for safety. Sixty—seven
`patients were evaluable for eflicacy (one patient had less than 6 weeks
`Ll antibody MPDL3280A, a systemic cancer immunotherapy, for
`follow up and therefore no efficacy evaluation). Of the efficacy—evaluable
`the treatment of metastatic UBC. MPDL3280A is a high—affinity engi—
`patients, 12 (18%) had tumours scored as PD—Ll IHCO. 23 (34%) as IHC
`neered human anti-PD-Ll monoclonal immunoglobulin—Gl antibody
`1, 2t] (30%) as IHC 2, 10 (15%) asIHC 3, and 2 (3%) as unknown based
`that inhibits the interaction of PD~Ll with PD—l (PDCD 1) and 37.1
`on tumour—infiltrating immune cells (see Methods for precise defini-
`(CD80)9. Because PD-Ll is expressed on activated T cells, MPDL3 280A
`tions). One patient had a PD—Ll IHC score of 2 or 3 for both tumour-
`was engineered with a modification in the Fc domain that eliminates
`inftltrating immune cells and tumour cells. Twenty-one patients with
`antibody-dependent cellular cytotoxicity at clinically relevant doses
`PD—Ll IHC 2 or 3 scores were enrolled before the cohort was expanded
`to prevent the depletion of T cells expressing PD-Ll. Here we show
`that MPDLSZSOA has noteworthy activity in metastatic UBC. Res-
`to include patients regardless of IHC status. Patients were pretreated
`ponses were often rapid, with many occurring at the time of the first
`with 62 (93%) receiving previous cisplatin- or carboplatin-based chemo—
`therapy (53 (79%) received previous cisplatin) and 48 (72%) receiving 2
`response assessment (6 weeks) and nearly all were ongoing at the data
`or more previous systemic treatments. Furthermore, many patients had
`cutoff. This phase I expansion study, with an adaptive design that
`poor prognostic factors at baseline 1 “2. including 50 (75%) with visceral
`allowed for biomarker-positive enriched cohorts, demonstrated that
`metastases, 12 (19%) with haemoglobin levels less than 10 gdl'" ', 22
`tumours expressing PD-Ll-positive tumour-infiltrating immune cells
`(33%) with creatinine clearance less than 60 ml min—', 39 (59%) with
`had particularly high response rates. Moreover, owing to the favour-
`an Eastern Cooperative Oncology Group (ECOG) performance score
`able toxicity profile. including a lack of renal toxicity, patients with
`of 1, and 26 (42%) whose time from previous chemotherapy was 3 months
`UBC, who are often older and have a higher incidence of renal impair—
`or less (Table 1).
`ment, may be better able to tolerate MPDL3280A versus chemother—
`In the safety-evaluable population, patients with UBC received
`apy. These results suggest that MPDL3280A may have an important
`MPDL3280A for a median duration of 65 days (range: 1—259 clays).
`role in treating UBC—the drug received breakthrough designation
`0fthese patients, 57% reported a treatment—related adverse event (AE)
`status by the US Food and Drug Administration (FDA) in lune 2014.
`of any grade, and 4% reported a grade 3 treatment—related AE, which
`We report on the safety and activity of MPDL3280A in patients with
`UBC who were enrolled in a UBC expansion cohort of a large phase I
`included one occurrence each of asthenia, thrombocytopaenia and de-
`creased blood phosphorus (Table 2 and Extended Data Table 1). There
`trial with an adaptive design. This progressive design has been used prev-
`were no grade 4 or 5 treatment-related AEs. Most treatment- related ABS
`iously to investigate immune checkpoint inhibitors in a spectrum of
`were grade 1 or 2, and many were transient in nature. Overall, decreased
`tumours and has resulted in regulatory approval in other settings” (see
`also http:llwwwspecialtypharmajournal.coml‘medical—newsloncologyl’
`appetite (grade ll2, 22%; grade 3M. 0%) and fatigue (grade ll2, 18%;
`grade 31%, 0%) were the most commonly reported toxicities and are
`51 l9—japanese« regulators—approve-the—first—pd— 1 —drug— f0 r—treatinent—
`thought to be related to immune system activationlJ (Extended Data
`of—melanorna). This UBC cohort was initially selected by PD—Ll immuno—
`Table 2). No investigator—assessed irru'ntme—related toxicities were reported.
`histochemistry (IHC) on tumour—infiltrating immune cells to test the
`hypothesis that PD—Ll—positive patients might specifically respond to
`For patients with a minimum of 6 weeks of follow-up, objective res—
`ponse rates (ORRs) were 43% (13 of 30; 95% confidence interval (CI):
`MPDL3280A. The cohort was subsequently expanded to include patients
`26—63%) for those with lHC 213 tumours and 11% (4 of 35; 95% CI:
`regardless ofFD-Ll status to determine whether PD- 1.1 negative patients
`4—26%) for those with IHC 0 0r 1 (011) tumours. The IHC 23 ORR
`could also respond. Overall, 205 patients were pre~screened and speci—
`
`mens were centrally analysed for PD—Ll expression. Of the available included a 7% complete response rate (2 of 30) (Figs 1c and 2). Among
`1Baits Cancer institute. Queen Mary University of Lo ndon. Ba rts Experimental Cancer Medicine Centre. London EC1M SEQ. UK. afale Cancer Center. 333 Cedar Street, WWW2] I. New Haven. Connecticut
`06520. USA 3Genenterh. inc. 1 DNA Way. South San Francisco, California 94080, USA. “Comprehensive Cancer Centers of Nevada. 3730 8. Eastern Avenue, Las Vegas. Nevada 89169. USPL 5Gustave
`Roussy. 114 Rue Edouard Vaillant. 94305 Villeiuif. France. Wall d‘Hebron Institute or Oncology (VHIO) and Will d'Hebron Universily Hospital. Passeig Vall d‘Hebron. 1 19-129. 08035. Barcelona. Spain.
`Ftilladoer Cancer Ce nter, Dana-FarberrBrigharn and Women's CancerCenter, Harvard Medical School. 450 Brookline Avenue. Boston. Massachusells 02215,USA.3Sarah Cannon Research Institute. 33 22
`West End Avenue. Suite 900. Nashville. Tennessee 32203. USA. 9University of Nevada School or Medicine and US Oncolognyorn preh ensive Cancer Centers of Nevada. 3?30 3. Eastern Avenue. Las Vegas.
`Nevada 89169. USA.
`
`558 | NATURE | VOL 5l5 I
`
`'2? NOVEMBER 201-1
`
`

`

`Hm
`
`
`Figure l i PD—Ll prevalence and response rates
`
`PD—Li prevalence in UBC tumours by IHC
`in patients with UBC. a. PD—Ll prevalence by
`immunohistochemistry (IHC) in patients screened
`PD-L1-positive
`for the PCD4939g clinical trial. PD—Ll positivity
`tumour-infiltrating immune cells
`PD-L1 -positive tumour cells
`was defined as 25% oftumour~infiltrating
`(no. of specimens (‘16))
`n = 205
`("0' Of spectmens (‘36))
`immune cells or tumour cells staining for PD-Ll
`IHC 3
`14 i?)
`by IHC. b, Representative images (X 20
`IHC 2
`3 (4)
`magnification) of I’D-Ll IHC staining of tumours
`IHC ‘l
`3? (18)
`from patients with UBC. 1:. Response rates,
`HO 0
`14s (71}
`including overall response, stable disease and
`progressive disease by tumour-infiltrating immune
`cell PD—Ll IHC status. Includes both confirmed
`and unconfirmed responses per RECIST v1.1.
`Five of seventeen responses were unconfirmed.
`Best response is not known for seven patients.
`
`18 (9)
`3? (18}
`89 (43)
`51 (30]
`
`
`
`Tumour-infiltrating immune cells
`Tumour cells
`
`
`c
`
`Tumour-infiltrating immune cells and objective response rates
`
`Objective response rate
`Stable disease
`Progressive disease
`
`:1 (as)
`n on
`n (961
`13 4 .3
`(95% Ci:(2:.5)-62.6)
`(95% 35222413,
`a 40.0
`(95% ell 204433)
`(9596:91'326 3)
`3 13.0
`(95% cl: 3.71—31 .r)
`
`IHC are (n = so)
`IHC 3 (n = 10}
`IHC 2 (n = 20)
`IHC 0n (n = 35)
`IHC 1 (n = 23)
`
`a {26.7}
`2 (20.0)
`6 (30.0)
`13 (3?.1)
`a (34.8)
`
`a (26.?)
`3 (30.0)
`5 (25.0}
`13 (37.1}
`s (34.8}
`
`IHC 0 (n =12}
`
`1 (3‘3)
`(95% Cl: 04—343)
`
`5 (41.7)
`
`5 {41.7}
`
`patients with IHC 2! 3 tumours and a minimum of 12 weeks of follow-
`up, an ORR of 52% (13 of 25; 95% CI: 32—70%) was achieved. Sixteen of
`the seventeen responders had ongoing responses, and all seventeen res-
`ponders continued on treatment with MPDLSZSOA at the data cutoff
`One patient who initially responded at the first response assessment
`later presented with new lesions, including a bladder mass thought to
`be consistent with pseudoprogtession. A biopsy of the new mass revealed
`extensive necrosis. This patient continued on treatment and had com—
`pleted 12 cycles at the time of the data cutoff.
`While the median has not been reached, duration of response ranged
`from 0.1+ to 30.3 + weeks for patients with IHC 21(3 tumours and from
`0.1+ to 6.0+ weeks for patients with IHC 01'1 tumours. Furthermore, while
`response to MPDL3280A was associated with the tumour-infiltrating
`immune cell IHC scores (P = 0.026), there did not appear to be an asso-
`ciation with tumour cell IHC scores (P = 0.93; Extended Data Table 3).
`Exploratory subgroup analyses demonstrated that IHC 2:3 and IHC
`Oil patients with an ECOG performance Score of 1 had ORRs of 33% (5
`of 15) and 14% (3 of 22), respectively, while patients whose time from
`previous chemotherapy was 53 months had ORRs of 33% (3 of 9) and
`19% (3 of 16), respectively. The ORRs for patients with visceral meta—
`stases at baseline was 21% (4 of 19) and 10% (3 of 29) for IHC 213 and
`lHC 0!] patients, respectively, while the ORRs for patients with no
`visceral metastases at baseline were 82% (9 of 11) and 17% (I of 6) for
`
`MC 213 and IHC Oil patients, respectively. The ORRs in current!
`former smokers and never smokers were 25% (11 of 44) and 26% (6
`of 23), respectively. In total, most patients (55%) had a reduction in
`tumour burden as measured by Response Evaluation Criteria in Solid
`Tumours, version 1.1 (RECIST v1.1) (Fig. 21)). Overall, responses were
`rapid and occurred at a median of 42 days from starting treatment
`(Fig. 2c). Patients with IHC 21'?) tumours and IHC 011 tumours had
`a median follow up of 4.2 months (range: 1.1+ to 8.5 months) and
`2.7 months (range: 0.7+ to 3.6 months), respectively. Twenty—five patients
`(37%) had been discontinued from the study due to disease progression
`(n = 17), death (:1 = 4), lost to follow- up [n = 1), physician decision
`(it = 2), or patient decision (in = 1).
`Over the course of treatment with MPDL3280A, cytoldnes and cir—
`culating cells were monitored. Transient elevations in cytoki nes, includ—
`ing interleukin (ID-18 and interferon (IFN)-'y, were observed by cycle
`2 day 1. A similar dynamic profile was observed for proliferating CD8 +
`HLA-DR+Ki-67+ T cells (Extended Data Fig. 3), consistent with the
`MPDL3280A mechanism of action. These markers were altered in all
`
`patients treated and were not associated with response.
`There is an urgent need for efficacious and well-tolerated therapies
`in metastatic UBC, as even first-line chemotherapy is poorly tolerated
`in a large proportion of individualsws. The study results presented
`here demonstrate that not only can MPDL3280A treatment achieve
`2? NOVEMBER 2014|VOL SIS NATUREISS‘?
`
`

`

`L_ETTEH
`
`Table 1 | Baseline characteristics of efficacy-evaluable patients with UBC
`Characteristic
`PD-L] IHC 2f3
`PD-Ll IHC Or’l
`Eflicacyvevaluable patients
`(n = 30)
`(n = 35)
`(n = 6?)
`
`
`
`
`Age (years)
`Median
`Range
`Sex ("1 (970))
`Male
`ECOG PS (:1 (95))
`0
`1
`Smoking status (1) (95))
`Current!previous smoker
`Site of primary tumour (n (%))
`Bladder
`Renal pelvis
`Ureter
`Urethra
`Sites of metastases at baseline (n (95))
`Visceral
`Liver
`Prior treatments‘(n (9:3))
`Cystectomyr
`Chemotherapy
`Prior platinum
`CiSplatin
`Carboplatin
`22 Prior systemic regimens
`Prior 300
`53 months from last prior chemotherapy (n (93))
`Organ function (it (95))
`Alkaline phosphatase EULN
`CrCi <60 ml mi n '1
`Haemoglobin <: 10gdl'1
`PD-Ll IHC (n (%))
`12 (119)
`12 (34.3)
`0(0)
`0
`23 (34.3)
`23 (65.?)
`0 (O)
`l
`20 (29.9)
`0 (0)
`20 (66.?)
`2
`10 (14.9)
`0(0)
`10 (33.3)
`3
`
`0(0) 2 (3.0) 0(0)
`
`Unknown
`BCG. Elaciiie Calmet‘te-Guerin: CrCl, creatinine clearance; ULN. upper limit at normal. Two patients have unknown IHC status.
`'n=29.tn=35.:n=66.§n=31:||n-- 62:'n-' 30: fin—34: "'n= 65.
`
`66.5
`42~86
`
`25 (83.3)
`
`14 (48.3)*
`15 (51.?)*
`
`20 (66.7)
`
`2? (90.0)
`l (3.3)
`0
`2 (6.7)
`
`19 (63.3)
`9 (30.0)
`
`20 (66.?)
`29 (96.?)
`29 (96.?)
`28 (93.3)
`T (23.3)
`21 (20.0)
`6 (20.0)
`9 (310)"
`
`4 (13.3)
`7 (23.3)‘i
`2 (6.9)*
`
`63.0
`36—81
`
`21 (60.0)
`
`13 (3?. l )1‘
`22 (62.9)i
`
`22 (62.9)
`
`28 (80.0)
`3 (8.6)
`3 (8.6)
`l (2.9)
`
`29 (82.9)
`12 (34.3)
`
`11 (31.4)
`31 (88.6)
`31 (88.6)
`23 (65.?)
`15 (42.9)
`25 (71,4)
`5 (14.3)
`16 (51.6)§
`
`10 (28.6)
`13 (38.2)“
`9 (26.5)“
`
`65.0
`36-86
`
`48 (21.6)
`
`2 T (40.9):
`39 (59.1):
`
`44 (65.?)
`
`55 (82.1)
`4 (6.0)
`5 (I5)
`3 (4.5)
`
`50 (?4.6)
`22 (32.8)
`
`'
`
`32 (4?.8)
`62 (92.5)
`62 (92.5)
`53 (79.1)
`23 (34.3)
`48 ('r' 1.6)
`11 (16.4)
`26 (41.9)“
`
`16 (23.9)
`22 (33.3):
`12 (18.5)“
`
`high response rates, but also that the likelihood of response can be in—
`creased by determining the PD- Ll status of tumour-infiltrating immune
`cells. Previous biomarker analysis with immune check point inhibitors
`has focused on PD-Ll expression on tumour cells rather than tumour-
`infiltrating immune cells. The observation that expression of immune
`infiltrates on pie—treatment tissue—which can be far removed temporally,
`anatomically and biologically from the metastatic tumours—correlated
`
`Table 2 | Treatment-related adverse eventsoccurring in two or more
`
`patients (grade 1-2) or in one patient (grade 3—4)
`Treatment-related
`All grades (:1 (56))
`adverse events' (:1 = 68)
`
`Grade 3—4 (:1 (96))
`
`3 (4.4)
`39 (5 14)
`All
`0
`8 (l 1.8)
`Decreased appetite
`D
`8(11.8)
`Fatigue
`0
`801.8)
`Nausea
`0
`6 (8.8)
`Pyrexia
`1 (1.5)
`5 (l4)
`Asthenia
`0
`3 (4.4)
`Chills
`0
`3 (4.4)
`Influenza-like illness
`O
`3 (4.4)
`Lethargy
`0
`2 (2.9)
`Anaemia
`0
`2 (2.9)
`Arthralgia
`0
`2 (2.9)
`Bone pain
`0
`2 (2.9)
`Hyperthermia
`0
`2 (2.9)
`Pain
`0
`2 (2.9)
`Platelet count decrease
`0
`2 (2.9)
`Pruritus
`1 (1.5)
`2 (2.9)
`Thrombocytopaenia
`0
`2 (2.9)
`Vomiting
`1(1.5)
`1 (1.5)
`Blood phosphorus
`decrease
`‘ National Cancer Instilute Common Terminology Criteria for Adverse Events. version 4.0.
`
`560 | NATURE | Vol. 515 ! 2? NOVEMBER 2014
`
`with outcomes provides some insight into the underlying stability of
`immune—related tumour surveillance in UBC. For example, the tissues
`examined here were originally obtained between 0 and 10 years before
`cycle 1, day l , with most tissues being obtained within 4 years (Extended
`Data Fig. 1). The association of response to MPDL3280A with PD-Ll
`expression on tumour~infiltrating immune cells was also recently ob—
`served in lung cancer”.
`While cross—study comparisons are limited, the 43% (95% CI: 26—63%)
`response rate achieved here in patients with PD-Li IHC 21'3 tumours
`provides evidence of noteworthy clinical activity of MPDL3280A in
`patients with UBC and compares favourably with that previously seen
`with single-agent salvage regimens'6'2". In addition, patients with PD-
`Li IHC Oil tumours had a response rate of 11% (95% CI: 41—26%). con-
`sistent with historic response rates of 9—1 1% in randomized studies for
`patients with relapsed metastatic UBC”. Responses in this heavily pre—
`treated population were also rapid and occurred in patients with poor
`prognostic features.
`Chemotherapy is challenging to administer in patients with UBC who
`have a median age at diagnosis of 73 years and multiple co—morbidities23
`(see also http:(iseer.cancer.govistatfactsihmiliurmbhtml). Many patients
`forgo chemotherapy for metastatic disease due to the toxicity and the
`limited durable benefit, and only approximately 40% ofpatients receive
`second—line treatment”. Therefore. the safety results with MPDLSZSOA
`are also encouraging. The larger and longer safety experience in the
`overall phase I study further indicates that M?DL3280A is well toler—
`ated, with AF. rates lower than many of the standard second—line treat-
`ment options for metastatic UBC".
`'
`To gain a better understanding of how the immune system responds
`to MPDL3280A. the levels of the IL- 18 immunostimulatory cytokine
`and IFN—y, which is stimulated by IL-18, were examined over several
`
`

`

`a
`
`Baseline
`
`Post-cycle 2
`
`Post-cycle 8
`
`
`
`
`MaximumSLDreductionfrombaseline(96)
`
`b
`
`IOUgo
`30
`in
`so
`so
`
`1100
`
`I Inc (ICJD
`I IHC {“311
`I IHC {1012
`I ”'10 {ICl3
`I IHCllcluniu'Iown
`
`
`
`
`
`t ttl'Dtiflilt “olefin-[ti
`i r1 -
`'
`-
`£006 PS Till I
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`_ ‘I‘: , 4%.}
`visceral metastases . _ . .
`Uvermelaslases in; '- .
`_ .11.
`
`”Ill.
`H9 <IogaI-‘
`100
`90
`g!
`$3
`2
`50
`E
`LU g 50
`2 - no
`a .2 w
`0']
`as i3
`: E
`o
`O
`-‘10
`E E —20
`01 .—.
`:1 .5 -30
`0
`D)“ —60
`i s -50
`5
`~70
`6
`:33
`not)
`
`I:
`
`
`
`
`— Complete response
`
`—-— Partial response
`Stable disease
`Progressive disease
`
`
`
` A New lesions r r 1 I 1 r I r I I r r
`
`
`
`
`Q Discontinued
`I
`
`
`
`
`
`
`
`a4 ‘IEIS 126 14? 163 139 210 231 252 2?3 294
`66
`02142
`Time on study {clays}
`
`Figure 2 I MPDL3280A anti-tumour activity in patients with UBC.
`a, Example ofa tumour response in a 68—year—old male who was initially treated
`for bladder cancer between 2011 and 2012. He underwent transurethral
`resection of the bladder tumour and intravesical Bacille Calmette—Guerin for
`localized recurrent disease (GSpTic). Subsequently, cystectomy and adjuvant
`cispl