You are on page 1of 9


ACS :: American Cancer Society Responds to Changes to USPSTF Mammography Guide... Page 1 of3

Press Room
close ~

American Cancer Society Responds. to

Changes to USPSTF Mammography
The American Cancer Society guidelines will not change; annual

mammography recommended for women 40 and over

Atlanta 2009/11/16 -November 16, 2009-The UnitedStates Preventive

Services Task Force (USPSTF) today announced that it is changing its
guidelines for mammography and no longer recommends routine.
screening for women between the ages of 40 and49. Below is a
statement from Otis W. Brawley, M.D., chief medical officer, American
Cancer Society.

''The American Cancer Society continues to recommend annual screening

using mammography and clinical breast examination for all women
beginning at age 40. Our experts make this recommendation having
reviewed virtually all the same data reviewed by the USPSTF. but also
additional data that the USPSTF did not consider. Vllhen
recommendations are based on judgments about the balance of risks and
benefits. reasonable experts can look at the same data and reach different

"In 2003. an expert panel convened by the American Cancer Society

conducted an extensive review of the data available at the time. ,which
was not substantially different from the data included in the current
USPSTF review. Like the USPSTF, the Society'spanel found convincing
evidence that screening with mammography reduces breast cancer
mortality in women ages 40-74, with age-specific benefits varying.
depending on the results of i:1dividuai trials and which trials were
combined in meta-analyses. And like the USPSTF, the American Cancer
Soder! panel also found that mammography has limitations - some
women who are screened will have false alarms; some cancers will be
missed; and some women will undergo unnecessary treatment. These
limitations are somewhat greater in women in their forties compared with
women in their fifties, and somewhat greater in women in their fifties
compared with women in their sixties. We speCifically noted that the
overaH effectiveness of mammography increases with increasing age. But
the Iimitatio:'1s do not change the fact that breast cancer screening using
mammography starting at age 40 saves lives. "As someone who has long
been a critic of those overstating the benefits of screening, I use these
words advisedly: this is one screening test I recommend unequivocally,
and would recommend to any woman 40 and over, be she a patient. a
stranger, or a family member.

UThe USPSTF says that screening 1,339 women in their 50s to save one
life makes scree:'1ing worthwhile in that age group. Yet USPSTF also says
screening 1,904 women ages 40 to 49 in order to save one life is not
ACS :: American Cancer Society Responds to Changes to USPSTF Mammography Guide ... Page 2 ?f3 ,

worthwhile. The American Cancer Society feels that in both cases, the

lifesaving benefits of screening outweigh any potential harms. Surveys of

women show that they are aware of these limitations. and also place high

value on detecting breast cancer early.

"With its new recommendations, the USPSTF is essentially telling women

that mammography at age 40 to 49 saves lives; just not enough of them.

The task force says screening women in their 40s would reduce their risk

of death from breast cancer by 15 percent, just as it does for women in

their 50s. But because women in their 40s are at lower risk of the disease

than women 50 and above, the USPSTF says the actual number of lives

saved is not enough to recommend widespread screening. The most

recent data show us that approximately 17 percent of breast cancer

deaths occurred in women who were diagnosed in their 40s, and 22

percent occurred in women diagnosed in their 50s. Breast cancer is a

serious health problem facing adult women, and mammography is part of

our solution beginning at age 40 for average risk women.

''The American Cancer Society acknowledges the limitations of

mammography, and we remain committed to finding better tests, and
currently are funding a large study to improve the accuracy of
mammography. In fact, data show the technology used today is better
than that used in the studies in this review, and more modern studies
show that mammography is achieving better results than those achieved
in these early experimental studies that go back as far as the mid-60's.
And as scientists work to make mammography even more effective, the
American Cancer Society's medical staff and volunteer experts
overwhelmingly believe the benefits of screening women aged 40 to 49
outweigh its limitations."

The American Cancer Society combines an unyielding passion with nearly

a century of experience to save lives and end suffering from cancer. As a
global grassroots force of more than three million volunteers, we fight for
every birthday threatened by every cancer in every community. We save
lives by helping people stay well by preventing cancer or detecting it early;
helping people get well by being there for them during and after a cancer
diagnosis; by finding cures through investment in ground breaking
discovery; and by fighting back by rallying lawmakers to pass laws to
defeat cancer and by rallying communities worldwide to join the fight. As
the nation's largest non-governmental investor in cancer research,
contributing about $3.4 billion, we turn what we know about cancer into
what we do. As a result, more than 11 million people in America who have
had cancer and countless more who have avoided it will be celebrating
birthdays this year. To learn more about us or to get help. call us anytime,
day or night, at 1-800-227-2345 or visit

###!docroot/MEDicontentJMED_2_) x_Amcrican_Cancer_Society_Resp... 3/2/2010

r . . . .,


Screening for Breast Cancer: U.S'. Preventive Services Task Force

Recommendation Statement
u.s. Preventive Services Task Force"
!)escription: Update of the 2002 U.S. Preventive Servi:es Ta>k The USPSTF recommends biennial screening mammography
Force (USPSTf) recommendation statement on screening for breast for women between the ages of 50 and 74 years. (Grade 3
cancer in the general populatiOJ:l. recom mend ation)
The USPSTF condudes that the current evidence Is ln5uffident to
Methods: The USPSTF examined the evidence on the efficacy of 5 assess t'le additlonaf benefits and harms of screening nammogra
screening modalities in reducing mortality from breast cancer: film phy in women 75 years or older. (1 statement)
mammography, dinical breast exaJnination, breast setf-exarnination, The USPSTF concludes that the current evidence is jr!S:.lifieient to
digital mammography, and magnetic resonance imaging in order to a>s~ the additional benefits and harms of dinical breast examina
wpdate the 2002 recommendation. To accomplish this update, the tbn beyond screening ma.'ilmography in womer. 40 years or oider.
USPSTF commissioned 2 studies: 1) a targeted systematic: evidence. (I statement)
review of 6 selected questions relating to benefits and harrns of The USPSTF recommends against clinicians teaching women how
screening. and 2) a decision analysis that used population to perform breast self-examination. (Grade D recommendation)
modeling techniques to compare the expected health out The U5PSTF condudes that the current evidence is insufficient to
comes and res:;Jurce requirements of starting and ending assess additionaJ benefits and harrns of either marnr.lography
mammography screening at different ages and using annual or magnetic resona.'1ce imaging instead of film mammography as
versus biennial screening intervals. screening modalities for breast cancer. (I statement)

Recommenci;:.tions: The USPSTF recommends against routine

screening mammography in women aged 40 to 49 years. The
decision to start regular, biennial screening ma:nmog~aphy before Ann Intem Meet. 2009;151;716-726.
the age of 50 years should be an individual one and take into For author affiliation, see end of text
account patient context, including the patient's values regarding For a list of the membe!:S of th: U5PSTF, see the Ap?endix (available at
specific benefits and ha.'Tns. (Grade C recommendation) wvJW.anna.!

Tht U.S. h(vmtivt Sm!iw Task Fora (USPSTF) Tht USPSTF rtcognias that clinical or policy tkcisiorJ
I m/:dus rtcommmdatjorIJ abow prromtivt carf srrvices involvf mOl( considmltiorJ t.han this body cf roi:/n,.C( aionf.
for jJl1timt! without ruogni:ud signs or symptoms of rhr Clinicians and poliC;'maluTJ should undmt:md thr evidmu
rargft condition. . but individwzbu dtcisitm making tc the spuific ptltimt or
It basu it! rccommmdtZtions on a -')'It/matic re'Jiew of fht sir-.uztio n.
roidmu ofthf bmtifits and hanm and an 4!SfJ.rmmr of fiu lut
bmfjit of tht str'"ict. SUMMARY OF RECOMIv',ENDATIONS A!~D EViDENCE
The USPSTF recommends against rourine screening
m;unmogra.phy in women :lged 40 to 49 years. The deci
See a/50: siOl1 to Start regular, biennial screening mammography be
fore the age of 50 years should be an individu:tl one and

take patient conrexr into account, inciudi"lg the patient's

Editorial comment ................. _ .... : ..' 750

values reg-...rding sp!:cinc bendits and harms. This is a C

Related articles ................... , 703, 727; 738

Summary for Patients.. : ...... , ............. 1-.44

The USPSTF recommends biennial screenir.g m2I'.l

mography for women aged 50 to 74 years. This is a B
Cfl\E quiz -- . The USPSTF concludt:S that the current e,,"iCence is
. Conversion of graphJc:s into slides
insufficient to assess the additional benefits a."ld hums of
Downloadab'e rec6mm~ndatio:1 summary
screeni."1g mammography in women 75 years or older. THs
is an I statement.

716/17 November l0091,:.nn.J., "f In:""",: M<:didn, Volume 151 Nomb", 10
,\L GUll}l1 !K1S

The USPSTF recommend, against teaching breast biopsies in women wirhour cancer, and inconvenience due
stlf-examination (ESE). This is a D recommendation. to false-positive screening results. Furthermore, one must
The USPSTF conch.:des that the current evidence is in also consider the harms associated with treatment of cancer
;ufficicnr [Q d.'ise:;s the addilional benefit.> and harms of clinical that would not become clinicaJJy apparent during a wom
hrc:m examination reEE) beyond s.:-reening mammography an's lifetime (o\'etdiagnosis), as well as the harms of unnec
ill women 40 years or older. This is an I statement. essary earlier rreatmt'nt of breast cancer that would have
The USPSTf concludes that the current evidence is become clinically apparent but would nor have shortened a
insuf'hcier:t to asses:; the additiollal benefits and harm:; or woman's life. Radiation exposure (from radiologic tests),
either digital mammography or magnetic resonance imag although a minor concern, is also a consideration.
in~ (MRl) instead or film mammography a, screenint!, mo Adequate evidence suggests that the overall harms as
dalities for breast cancer. This is an I statement. sociated with mammography are moderate for every' age
Figures 1 a.nd 2 summarize the recommendations and group considered, although the main components of the
provide suggesrions for clinical practice. harms shife over rime. Although false-positive test results,
Table 1 describes the USPSTf grades and Table 2 overdiagnosis, and unnecessary earlier treatment are prob
describes the USPSTF classiilcarion of levels of cen:linry lems for all age groups, [llse-posirive results are more com
about net benefir. mon for women aged 40 to 49 years, whereas overdiagnosis
See Clinical Considerarions for specific popularions at is a greater concern for women in the older age groups.
incrca;,ed risk and for suggestions f1Jr practice regarding the There is adequate evidence that reacl1ing BSE is asso
1 statements. ciated with harms that are at least small. There is inade
quate evidence concerning harms of CEE.

USPSTF Assessment
Importance The USPSTF has reached the following conclusions:
Breast cancer is [he second-leading cause of cancer For biennial screening mammography in women aged
death among women in rhe United States. \XTidesprcad use 40 to 49 years, there is moderate certainty that cile net
of scrcl:ning, along with treatment advances in recent years, benefit i, small. Ahhough the USPSTF recognizes that the
have bem credi:ed wich significant reducrions in breast benefit of screening seems equivalent for women aged 40
cancer morrality. to 49 years and 50 to 59 years, the incidence: of breast
cancer and the consequences differ. The USPSTF empha'
sizes the adverse consequences for most women-who will
i\1ammography, a..s well as physical examination of the
not develop breast cancer-and rherefore use the number
breasts (CBE ar.d ESE), can decen presymptomatic breast
needed (Q screen (Q save 1 life as irs metric. By this metric,
cancer. Because of its demonstrated effectiveness in ran
the USPSTF concludes that there is moderate evidence
ciomi7,ed, comrolled trials of screening, film mammogra
that the net benefit is small for women aged 40 ro 49 years.
phy is the standard for detecting breast cancer; in 2002, the
For biennial screening mammography in women aged
USPSTF found convincing evidence of irs adequate sensi
50 ro 74 years. there is moderate cenain)' that the net
civic}' and ;pt'cif:ciry.
benefir is moderate.
Benefits of Detection and Early Intervention For screening mammography in women 75 years or
There is cOllvincir.g evidence thar screening with film older, evidence is lacking and rhe balance of bench ts and
mammography reduces breast cancer mortality, with a h(lrms cannot be determined.
greater absolure reduction for women aged 50 to 74 years For the teaching of BSE, rhere is moderate certainty
than For women aged 40 to 49 years. The strongest evi that the harms outweigh the benefits.
del ICe for rhe greatest benefit is among women aged GO ro For CBE as a supplement to mammography, evidence
69 years. is lacking and the balance of benefits and harms cannor be
Among women 75 years or older, evidence of benefits determined.
of mammography is Jacking. for digital mammography and MRl as a replacement
Adequate evidence suggem that reaching ESE does for mammography, the e\ridence is lacking and the balance
not reduce breasr cancer morraliry. of benefits and harms cannor be determined.
The evidence for additional effects of CBE beyond
r:1ammography on breast cancer morrality is inadequate.
The evidence for benefits of digital mammography CWHeAL CONSlDERAT!or~S
.1:d :\1R! of dle brca~t, as a mbsr:tutc for iilm mamm0f,
Patient Population Under Consideration
taplly, is also lacking.
This recommendation statement applies to women 40
Harms of Detection and Early Intervention y[:ars or older who :ire nor at increased risk for breast can
The harms resultir.g from screening for breast cancer cer by virtue of a known underlying genetic mutation or a
include psychological harms, unnecessary imaging rc:sts and hisrory of chest r;;.ciiarion.
Assessment of Risk meir patients by considering the evidence in favor of a
Increasing age is the most important risk faeror for Structured, standardized examination (2).
breast cancer for most women. Women wiiliout known
deleterious genetic mutations (such as BRC-41 or BRO.2)
Digital A1ammography
may still have oilier de.o.ographic, physical, or hisrorical
Potential Prevtnt4.bfr Burden. Digital ma..'T..mography
risk factors for breast cancer, but none conveys a clinically
derecrs some cases of cancer nor identified by film mam
important absolute increased risk for cancer.
mography; film mammography detec-..s some of can
Screening Tests cer not identified by digital mammography. Overail dmc
In recent decades, the early detection of breast cancer tion is similar for many women. For wo~en who are
has been accomplished by physical examin.ation by a clini younger than 50 years or have dense breast tissue, overall
cia.n (CBE), by a woman herself (BSE), or by ma.tnmogra derection is somevvnat higher wirh digital mammography.
ph)'. Standardization of mammography practices enacted It is not clear whether this addicional detection would lead
by the Mammography Quality Standards Act have led to to reduced mortality from cancer.
improved ma,:nmography quality. Clinicia."lS should refer Potential Harms. The possibiliry of false-positive test
patients ro Mammography Quality Stan.dards Act-cerrid resufrs is similar for Elm a.."ld digital mA..'n.rnography. 1t is
facilities, a listing of which is available at \ uncer-..ain whether overd1a"GIlosis occurs more wrt:.1:! digital
Icdrb!ma..rnmography/cerrified.html. mammography than wit.1:! film mammography.
COltS. Digital mam.-nography is more expe.ruive than
Screening Intervals
Elm mammography.
In trials dut dem.onstrated the effectiveness of mam . Current Practice. Some clinical practices are now
mography in decreasing breast cancer monaliry, scree.wg switching. thcir mammography equipment from file to
was performed every 12 to 33 months. The evidence re digital. This may cunail the availability of Elm mammog
viev...ed by the USPSTF indicates that a large propomon of raphy in some areas.
the beneiic of screening mammography is maintained by
bier.nial screening; and cha.!"lging from ar...nual to biennial
screening is likely to reduce the harms of mammography Magnetir:: Resonance Imaging
screening by nearly half. At the same rime, beneht may be Potmtial Pl'lvmtabk Burdm. Studies of the use of
reduced when e..:tendlo:g the interval b~'ond 24 months; contr.lSt-enhanced l\oo for breast canca: screening bye
therefore the USPSTF recommends biennial screening. been conducted orJy in very high-risk populacions .. In
these smdies, MR1 detec.:ed more cases of cancer man did
Treatment man::mograpby. It is unknown ",-hether detecting t..~ese ad
Effective treatmems, including radiacion, chemother ditional cases of cancer would lead to redued breast cancer
apy (inclurur:g hormonal treatment), and surgery, :lIe avail monallty.
able for invasive carcinoma. Although tIle standard treat Pounti~i Harms. MRI :eqcires t..l:le
mencs women receive for ductal carcinoma in sim (DCIS) injeccion of contrast material. Studies of MRI screening
include surgical approaches as weU as radiation ,Uld hor have shov.'n that MRI yields many more false-positive re
monal therapy, considerable debate exists about the opti sdts than does ma.mmography. Magnetic resonance imag
mal treatment strategy for this condition. ing has the potential to be associated wit.~ a gr::ater degree
Considerations for Practice Regarding I Statements of overdiagnosis t..~an ma:m.rcography.
Clinical Breast Examination COlt!. Magnetic resonance imaging is much more ex
Pour-tial Prevmtable Burdm. The evidence for CSE, pensive u~an cirher film or digital ma..'TImogrnphy.
although indirect, suggeStS that CBE may detect a substan Current Practice. l\,14gnetic resona.."1.ce imaging is not
tial proportion of cases of if it is the o:Jy screening currently used for screening women at average risk for
test available. 1r1 parts of the world where mammography is breast cancer.
infeasible or unavailable (such as India), CBE is being
investigated in this VI"ay. 5creening Mammography in Women 75 Years or Older
Potential Harms. The potencial harms of CBE are Pountid Prromtabfr Burden. No women 75 years or
u':!ought to be smill but include filie-positive test results, have been included in tb.e multiple randomized clin
which lead to anxiety and breast cancer worry, as ical trials of br::ast cancer screening. Breast cancer is a lead
repeated visits and unwarraIlted imagi..t"lg and biopsies. ing cause of death in older women, ,vhich rnight suggest
Com. The principal cost of CBE is the oppOrtur.ity that the benefits of screen.i..."lg could be i:nportant at t:.llls
cost incurrd by cli..'1icians i.'l the patient enCoULter. age. However, 3 facts su.ggesr that benefits from screening
Cumr.! PrlUtice. Surveys suggest (l) that the CBE would probably be smaller for this age group ::han for
techl'ique used in the United States currently laoo a W ...!l women aged 60 to 69 years and probably deCrease with
dardapproach and reporting srancards. Clinic:ar:.s who are increasing age:: 1) the benefics of screening occu.r only sev
committed to spending the time on CBE would benefit eral years mer uie actual screening test, whereas the per-
718 rli Ne,,,mber 2009 Ir""d, af!ne",""", '.{odicit;, jVolume 151' Numb'l lC v,ww.annals.or
Cf'nr;lE,C of women "dlO SUITive IO:1g enough to benciit de s),stematic re\'iew (7), whicll in(:orporates a new random
creases with age; 2) a higher percentage of the eype of ized, controlled trial that estimates the "number needed to
breasl cancer detected in this age group is the more easily icvite for screening to extend one woman's life" as 1904
treated estrogen n;:cepror-posirivc type; and 3) wumen of for women aged 40 to 49 years and 1339 for women
1his age are at much greater risk for dying of other cOI!di 50 to 59 years. Although the relative risk reduction is
(ions that wodd not be affected by breast cancer screening. nearly identical (15% and 14%) for these 2 age groups, the
Porenri,1! Harms. Screening detects not only cancer risk for breast cancer increases steeply with age starring at
that could lead to a woman's death but also cancer that will age 40 years. Thus, the absolute risk reduction from
lint shonell a \NOman's life. \X'omeJJ cannot bellefit from screening (as shown by the number needed to inyite to
SUl can be harmed hv-the , , and treatment of this
disClwerv screen) is greater for womell 50 to 59 ycars rhan for
second type of cancer, which includes both cancer lhat those aged 40 to 49 years.
might some become clinically app;!reIH and cancer that The currenr USPSTF statement is also informed by
cever will. Detection of cancer that \vould never han: be the Cancer Intervention and Surveillance !'.10dcling >Jet
come clinically apparent is called ()Z'erdi,tgnosis, and It is work (ClSNET) modeling studies (8) that accompany this
L:slIally followed by OVl;'flrearmenL Bt:causc: of a shortened recommendation. The Task Force considered both "mor
life span among women 75 years or older, the probability tality" and "life-years gained" OUlcomes. III this case, given
o~ overdiagnosis and unnecessary earlier rrcaunem in that the age groups (40 to 49 years and 50 to 59 years) are
creases dramatically after abom age 70 or 75 years. Over adjacent, (he Task Forct e1encd to emphasize the mortality
diagnosis and unnecessary earlier treatment arc importa11l ourcomes from the modeling studies. Of the 8 screening
potential harms from screening women in this age group. strategies found most efficient, 6 Start at age 50 years rather
Currellt Pra(fict. Studies show char many women 75 than age 40 years. The frontier CUf\'es for thc mortality
rears or older ar(;' currently bt"ing screcntd. outcome show only small gains but larger numbers of
Useful Resources
mammograms required when screening is started at age 40
years \'ersus age 50 years.
Other USPSTF recommendarions on screening for ge
netic susceptibility for breasr cancer and chcmoprevenrion
In conclusion, the liSPSTF reasoned that rhe addi
rional benefit gained by narting screening at age 40 years
of breast cancer are available on the Agellc}' for Healtheare
rather than at age 50 years is small, and that moderate
Research and Quality \Veb site (w\\"w.prevemlveser\,ices
harms from screening remain at any age. This leads to the
C recommelldation. The USPSTF notes that a "C" grade
is a recommendation against routine screening of women
OTHER (ONSiDERATiONS aged 40 to 49 years. The Task Force encourages individu
Implementation alized, informed decision maleing about when to start
The Task Force on Community Preventive Sef\'ices mammography screening.
h::s reviewed the evidence on methods to increase breast
cancer including reminder systems and ocher in-

Explanation of Change in Recommendation

A series of randomized clinical trials that would com
pare the results of SlOpping brea$t cancer screening at dif
The 2002 USPSTF issued a B recommendation for
ferem ages (by hrsr comparing stopping scre(;'ning at age 75
screening mammography tor women 40 years or older.
rears with continued screening, and then further compar
However, it went on [Q say:
ing stopping screening at earlier ages, depending on the
The precise age at wbich the hme5:5 from SCI ecning
n:,uhs of the first study) would be ethical and informative.
marnmogr;;pl1y jU5tifl' the ?Drcnriai harms is a subjec Extended follow-up of this type of study might also
tive and should take inca account patient provide useful information about overdiagnosis in this age
preference,. Ciniciam ,hould infol:n women ;:.bout UiC group. In general, more swdies of ovcrdiagnosis, including
potencial benefits (reduced chance Hf dying fwm brcasr comparisons of liferime breast cancer incidence among
cancer), potential harms ([or exampl::, re similar screened and unscreened 'women, \vould be helpful.
suhs, unrH:cessary hiopsies), 2nd limitations of the test Studies on overdiagnosis might also include long-rerm
th:n ~P?:Y to women their age. Clinici"ns should teI! [ollow-up of women with probabJe missed cases of DClS
women tb.t the bal~nce of benefits and potencial hJfInS on the ba.:;is of microcalcihcations th:H were missed in an
lit" m"mmogral'h~' improves wiLl! age fur earlier mammogram. Such studies could provide the per
W('men between the ages of 40 and 70 (6)' centage of these women who develop invasive breast cancer
o\'er the next i 0 or more rears.
The updated l;SPSTf: lc'cOJ~uTI(':lld31ion endorses this Randomized clinical trials of film versus digital mam
to starr $crcening. Ho\\'en~r, [() deciding when mography among women with dense breast rissue, with
(he cum:nr LJSPSTF is now furehcr illformed by J new sufhcic;-)t [o;low-up to detect st;ige shifts (reductions ot
,j , ' j 1
eu N 1(.".L GUID ELI!''; ES Screening for Bre;.s;: Cancer

late~5tage cancer) or decreases in clinical int:::rval cases, In addition to the systematic review of scree:ling tests,
would :ilio be ethical and helpful. the USPSTF requested a report from the CISNET Breast
Better ll.' of cemin facets of ruInor biol~ Cancer Modding Group to provide data from comparaciye
ogy is needed. particula..rly how age, race. breast density. decision modds on opci.r::lal mning and stopping ages a.'1d
and other factors may predispose cer-.". lin wom::n towa..rd intervals for screming mammography (8).
tlL"I10rS with faster grow-.h rates and greater lethality. This
Accuracy of Screening Tests
would improve r.~e ability to deterll'Jne at diagnosis which
pacientS CQ.ll be treated minimally. Mamrnograpby, CBE. and BSE are recognized ap
proaches for breast cancer screening. Since the 2002
USPSTF recommendation statement. digital (as opposed
to Elm-based) mammography has been increasingly used,
and l\1RJ is being used with greater frequency for screeni.:lg
Burden of Disease
women at increased risk for breast cancer. The ser.siri"ity
Breast cancer is the most frequently diagnosed can of mammography screening is 77% to 95%, vvhe:eas spec
cer in women in the United States, not including skin ificity is 94% to 97% (16). Multiple factors, including age,
cancer, and is .second only to lung cancer as a cause of time sIDce last ex2lllination. breast tissue density, equip
cancer deaths. ln 2008, an estimated 182 460 cases of ment. and the skill of t.~e interpreting radiologisr can affect
invasiv.e ca.'1cer and 67 770 cases of in siru breast cancer sensitivity and specificity (17). A single, large compa..rison
were diagnosed and 40 480 breast cancer deaths oc . srudy of film and digital mammography (18) de.nor.strated
curred (9). The National Cancer Institute, on the basis similar diagnostic accuracy for the 2 methods, althougb.
of Surveillance Epidemiology and End Result data, es~ digital mammographyv.'a5 bener at detecting lesions in
tim ares the lifetime risk for a wom:an to develop breast women who were younger thw 50 years or premenopausal
cancer at 12% (10). The risk for breast cancer increases or had r:l.diographidly dense breasts. Studies of MRI in
with age. The 1O~year risk for breast cancer is 1 in 69 high-risk women without cancer (19) showed that Y..RI
for a WOman at age 40 years. 1 i11 42 at age 50 years, and has a sensitivity of 71 % to 100% and a specificity of B1%
1 in 29 at age 60 years (ll). The incidence rate of breast 1:0 97%; MRI is therefore recominended by the .American
cancer has increased since the 1970s; however, recent Cancer Society for women at high risk for breast cancer.
data show mat mis rate see.lns to be decreasing. bom However, no srudies have been done on using MPJ to
oyerall and on an age-adjusted basis. The incidence rate screen women at average risk.
in 2003 was 124.2 per 100 000 women, a 6,7% decrease Clinical breast aarnination has a sensitivity of 40% to
from the previ.ous year (12). Discontinuation of hor 69% a.'1d a specificity in ~e range of 88% to 99%. For
mone replacement therapy may be largely responsible BSE, sensitivity ranges from 12% to 41 %. lower tha.'1 that
for this observed decrease (12, 13), although slowed of CBE and mammography, wd is age-dependent (16).
gro"l1:h or even a decline in screening mammography
also may have contributed (14). Breast c:aneer mortality Effectiveness of Early Detection
has been decreasing since 1990 by 2.3% per year overall The nev,'ly updated meta-analysis by Nelson and col~
and by 3.3% for women aged 40 to 50 years. This leagues (7) connrms an earlier (16) t.~at screening
decrease is largely attributed to the combination of mammography reduces monality. Improvements in the
mammogr2.phy screening with improved treatment (15). relative risk (RR) for deam due co breast cancer for women
aged 39 to 49 years and 50 to 59 years are simil:u- at 0.85
Scope of Review (95% CT, 0.75 to 0.96) and 0.86 (CI, 0.75 to 0.99), re
The systematic evidence review undertaken in support spectively. AIl even greater improvement was found for
of this recommendation (7) addressed the: efficacy of 5 women aged 60 to 69 years (RR. 0.68 [CIt 0.54 to 0.87]).
breast cancer screening methods for reducing breast ca.'l.cer Results were uncertain for older women, v.ith a new report
mortality-film mammography, CBE, BSE, digital marI: from a previously included trial with longer follow-up data
mography, and MPJ- by usir.g published reports of r:an showing an RR of 1.12 (CI, 0.73 to 1.72) for breast cancer
domized, controlled screening r.:ials and specifically up~ de.ath associated with screening women aged 70 to 74 years
dated ir.formation from mammography trials a..'11ong (based on a .small number of participants). At t.~e ti.'11e of
women in the age groups of 40 to 49 years and 70 years or the previous meta-analysis. data trom 2 trials showed an
older. Informacion on harms of breast cancer screening, RR of 0.78 (CI, 0.62 to 0.99) for breast cancer death in
such as false-positive teSt results, pain, anxiety. and biopsy screened women aged 65 to 74 years (16). Mor..ality reduc
rates. were sought from multiple sources, including system tion a.'l.d life.-years gained (8) were both consicered L'TIpOr
atic reviews. meta-analyses, :and recently published litera tant outcomes of screer-ing in forming this recomme...'1da
tUle. To assess the follow-up testing a.'1d other outcomes of don; in me end. mortality reduction as. observed in trials
a mammography screening program, the re\'iewers in was me metric chosen w express tI:lI~ benefirs of scm:.!".ing
cluded data trom t.~e Breast Ca.ncer Surveillance Consor .in e41ch group. Digital mammography has been sho'i'.'TI
tium from 2000 to 2005. co perform similarly to film mammography, although it
Scre~ning fe.r Brea.'t Can cer I J"1(;\.1

has never been srudied in rcLltion 10 health Olllcomes. Clinical Breast Examination
~\'lagnetic resonance imaging has nor been evaluated for its A theoretical harm of eBc is lhat its lower specitci,y
potencial benefit in screening averilg,c-ri:;k women. Clini::al than that of mammography may result ill more women
breast examll1JtiOIl is inexpensi\'c and easy to perform; undergoing biopsy. Few data are available to evaluate
however, few studies have compared the cffccrivene$s of harms associated wilh eRE.
eRE with that of no screening and none has examincd
CBE and mammography versus mammography alone,
Breast SelfExamination
Two large trials of reaching BSE olmidc the United States
One study indicated that anxiety was not a concern
(7) del710flstraled no monaliry benefit in Ihe illtervendoll
with BSE. The 2 available trials (20, 21) indicated that
more additional imaging procedures and biopsies were
Potential Harms of Screening done for woml'n who performed BSE than for cOl1trol
Mammography partiClp:l11ts.
False-positive results are common with mammography
Estimate of Magnitude of Net Benefit
;lnd Ciin calise anxiety and lead to additional imaging stud
1n 2002, the USPSTF concluded that there was fajr
ies and il:vasive procedures (such a:; biopsy or line-lJeedle
evidence that mammography screening t"'ery 12 to 33
aspiration). Fahe-positive results ilnd accompanying addi
momhs could significantly reduce bre,lst carlcer mortality.
tional imaging studies are more common in younger
Tbe evidence was strongest for women aged 50 to 69 rears,
women. Biopsies may occur as a consequence of false
with weaker evidence supporting mammography screening
positive mammography results; biopsy rates arc more com
for women aged 40 (0 49 years. Since that recommenda
mon among older women. Anxiety, distress, and other
tion, I new trial and updated data from an older study
psychosocial effects can exist WiUl :abnormal mammogra
have been published thar specifically address screening in
phy results but fOf[unareiy are usually rransiem, and some
women in the younger age group. These findings were
research suggests that tbese effects are !lot a barrier to
combined in an updated meta-analysis, which resuJled in
screening, False-negarive results occur at a relatively low
anRR for breast cancer death of 0,85 (CI, 0.75 lO 0.96; 8
ratc for all ages, but are slightly higher in women older
trials) and a number needed to invite for screening of [904
than 70 years. Other potential harms, such as pain caused
by the procedure, exist bur are thought [0 have little dfect
(Cl, 929 to 6378) to prevent 1 breast cancer death in
women aged 39 to 49 years. A meta-analysis of 6 trials
on mammography use,
among women aged 50 to 59 years and 2 trials among
Overdiagl1osis can occur when screening detects early
women aged 60 lO 69 years provided pooled RRs for breast
stage invasive breast cancer or DClS in a woman, typically
cancer death in the screened group of 0,86 (CL 0.75 to
older, who is likely to die from another cause befnre the
0.99; number needed to in\'i1e, 1339 [CI, 312 to 7455])
blea.,t cancer would be clinically detected. Overdiagnosis
and 0.68 (Cr, 0.54 to 0,87; number needed to invite, 377
can J.lso occur in younger women if a detened DClS or
other early-stage lesion never progresses to im'asive cancer.
lei, 230 to 1050]), respectively. Only 1 smdy provided
data Oil women older than 70 years, yielding anRR of 1. J1
\1cthods for estjmating overdiagnosis at a population level
(C1. 0.73 to 1.72), although an RR of 0.78 (CI, 0.62 to
are nor well e:.;tablished, and thus the proportion of all
0,99) was found for women aged 65 to 74 vears b\' com
detecn:d DClS lesions that constitUte o\'erdiagnosis is un
bining the results of 2 studies ~used in the 2002 rc\;iew.
certain (7)_ Similarly, unnecessary earlier treatment can oc
A decision analysis performed for the USPSTF pro
cur at any age when screening detects a slower-growing
jecced that biennial screening produced 70% to 99% of the
cancer chat would have eventually become clinically appar
benefit of annllal screening, with a significant reduction in
ent but would never have caused d::ath.
the number of mammograms required and therefore a de
Ra,diatiofl exposure may illcrease the risk for breast
creased risk for harms. Screening between the ages of SO
cancer, but usually only at much higher doses than those
and 69 years produced a projected 17% (range, 15% to
used in mammography, although regular mammography
23%) reduction in morralit}' (compared with no scrcen
could contribute to cumulari\'e radiation doses from addi
ing). whereas eXlending the age range produced only minor
tional imaging [or other reasons (7).
improvements (additional 3% reduction from starting ar
age 40 years and iOt'h from extending to age 79 years) (8)_
Digital Mammography and Magnetic Resonance Imaging The USPSTF noted with moderate certainlY that the
Digiral mammography can be expected to have harms net benefit~ of screening mammography in wom;n aged 50
"imi!ar ro those of film mammography, No "tudies h3.\'c to 74 years were at least moderate, and that the greatest
evaluared MRI screening of average-risk women; in tht:Ory, benefi~ were seen in women abed 60 to 69 ye~s. For
because MRI could be considembly Ie~s specific than mam women aged !to to 49 years, the USPSTF had moderate
mography screening, it could potentially be associated with certainf}' that the net benefits were smaiL Because of d:e
higher biopsy rates and greater degrees of overdiagnosis if uncertainties related to harms of screening. particularly
t:sed in lew-risk port:l:l[ions. ovcrciiagnosis. and the near toral lack of trial data for older

women, the USPSTF had low cerratl1ty about

,.,.. L t
ems ot screen:ng mammograpr.y ,or women 75 years or Numerous organizations have provided b:east canc:::r
older. screening recommenda:ions. These recomme.nciacions are
How Does Evidence Fit With Biological Understanding? summariz.ed below..All recommendations are for women
Current knowledge about the development of breast not at increased risk for breast cancer. .
cancer is limited. The effectiveness or screening mam In 2003, the i\..."Deric211 C:l1lcer Society recom
mography seen in trials presumably results from the mended annual mammography beginnL."lg at doge 40
early detection of smaller, earlier-stage tumors, which years, annual CBE after ui.e age of 40 years (25). It does
are more responsive to available treatments. ,Although not recommend MPJ for women at average risk for
the roost common breast cancer occurs in the epithelial b;-east cancer and st;;.tes that there is insufficient evi
cdls that line the duct system of rhe gland (ductal car dence to recommend BSE.
cinoma), the sequence of development of invasive can The .American Merucalll.Ssociation, in 2002 (26), and
cer is not entirely known. For example, DCIS does not the National Comprehensive Cancer Network, in 2009
always represent a precursor to invasive ducral cancer. (27), have made recommendations similar to those of the
Studies of women with unueated DCIS showed pro American Cancer Society, except for the inclusion of a
gression to invasive disease in half or fewer of the cases. positive recommendation for BSE.
Because DCIS is orren found only by mammography, its The American Academy of Family Physicians has en
incidence has increased s:eadily since the advent of dorsed the USPSTF recor:::lJllendation on breast cancer
widespread screening mammography. In 1983. 4900 screening in the past (18). The American College of Phy
cases of DCIS were diagnosed; by 2008, that number sicians recommended in 2007 that screerung mammogra
was expected to be 67 770 (22). Because the likelihood phy decisions ill women aged 40 to 49 yea:s should be
that DCIS will progress to invasive cancer is unknown, b;:sed on individualiz.ed assessment of risk for breast can
surgical removal-with or without adjuvant treatmem ce.r; that clinicians should inform women aged 40 to 49
may represent overdiagnosls or oyertreatment. Lobular years about ,he potential benefits and harms of screening
carci.noma in situ, in coorrast, is not considered a true mammography; and that clinicians should base screening
precursor lesion but connotes a higher risk for subse ma.'Tlmography decisions on bendir.s and harr..5 of screen
quent invasive lobular or ducal cancer in either breast. ing, as well as on a woman's preferences and breast cancer
Lobular carcinoma in situ is often multifocal, appearing risk profile (29).
in several distinct locations. Knowledge of ...v hat deter In 2001, the Canadia.'1 TO?Sk Force on Preventive
mines the rapidity with which invasive cancer can Health Care recommended mm:unography every 1 to 2
spread (rcr::lor characteristics, host factors, hormonal years beginning at the :age of 40 years and recommended
trIggers) is limited (23). Because of these dements of CBE as part of a periodic evaluation (eyery 1 to 3 years) for
biological uncert;linty, it is dear that lesion sensitivity women aged SO to 69 years (30). It does not recommend
alene is not a su.ffi:::ient metric for ass::ssing effectiveness BSE.
of new screeningn'iethods (24). In 2003, the .American College of Obs:e::rics and Gy
necology recommended ma:r.mography every I to 2 years
for women aged 40 to 49 years and a.'1nualJy after th.e age
UPDATE OF PREVIOUS USPSTF RECOflllMENDATION of 50 years (31). It recommended CBE for all women. a..,d
This recommendll.uon updates the 2002 recommenda Dored that BSE can be recoI!4-ncoded.
tion by providing specific recomme-'1dations for mamr.:lOg In 1009, the World Health Or~anization recom
raphy screening by age. The previol!s recommendation mended mammography every 1 ro 2 yea7:s for women aged
statement recommended screening mammography every 1 50 to 69 years, but docs not recom:nend CBE or BSE (32).
to 2 years for all women older tha.'1 40 years. The USPSTF
now recommends against !Curine screening of women aged From th~ U.S. Prcv~nciv~ Services T;sk Force, Agency for He..I6=~
40 to 49 rars (C recommendarion), recommends biennial ~.earcb and Qu:ility. Roci..."illc, M:1..'yiL'ld.

screening r.t:unmography for all women aged 50 to 74

years (B recommendation), ~.,d provi.des an I sratement Disclaimer: Recor:um:ndltcions m"d~ by th~ USPSTF are indc:pc..'ldent of
regarding screening of women oider than 75 years. The t.he U.S. gove:nment. They should nor be c:onscnw::! IlS ~n o:nci:J pmi ..
USPSTF now recommends against teaching ESE (D rec ~ion of the Agen:y for Heal ::hare Rcse;rch 2nd Qu~ity 0 r the U.S.
Dep:lI"CTlenc of H~::h and Hwn:rn Services. .
ommendation), replacing ,he previous statemer..t of insuf
ficient evidence:. Tne evidence for continues to be
assessed as insufficient. Digital maJ'nmography a.'1d MRI as Financi;:l Support: The USPSTF is an i.ndepend~l, yoiu..::.'Il'Y bod)'.
. tools were not addressed in the 2002 recommen The U.S. Congrcss mwda:6 that the Agency for neol::h=e Research
dation m.ternem; the USPSTF concludes t,l,.at the evidence ;,nd Qu:ali!}' suppOrt the Ope::lDons of the USPSTF.
is insuFficient to assess the burns or bc:ne:ir.s of (.i.esc: rnee.h
ods for screening. Potential Conflim of interest; None disclosed.

722117 "c"~rnbc, 20091 f~m::J, nfl"'<["~i \bJi.:in< 1Volume l51 Numb", 10 WI'IW.lnnaI1.Or~