BMJ 2012;345:e7536 doi: 10.1136/bmj.

e7536 (Published 16 November 2012) Page 2 of 10

RESEARCH

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 1 of 10

R esear ch

diagnosed within the same period but before they had been invited to mammography screening. We investigated whether breast cancers detected in the interval after a normal mammography screening result but before the next scheduled screening are more lethal and thus may need more aggressive treatment than non-screen detected breast cancers.

of women with these tumours might be worse than for those whose breast cancer is overlooked, and with increasing screening rounds and more experience the prognosis would be worse. We therefore carried out a secondary analysis restricted to interval breast cancers stratified by screening round.

RESEARCH

Study population
From the cancer registry database we retrieved information on date of diagnosis, age at diagnosis, county of residency, classification of the canceraccording to the pathological tumour, node, metastases (pTNM) classification (International Union Against Cancer guidelines), 21 and tumour stage in all women with a first diagnosis of invasive breast cancer at age 50 to 72 years in Norway between 1 January 1996 and 31 December 2006. Thestage ofbreast tumour is coded as I (localised cancer), II (regional cancer), III (cancer fixed to the skin or the chest wall), or IV (cancer with distant metastases). To determine whether a diagnosis was made before or after invitation to 1 4 123 , Rulla M Tamimi assistant , screening we linked data on all the women toprofessor the screening 16 professor database at the cancer registry. We then further categorised women with breast cancer diagnosed after invitations to screening as interval cancers if they met our criteria. 2 University of Oslo, Faculty of Medicine, Institute ofHealth and Society,

Methods
Breast cancer screening programme
Since 1951, reporting of cancer diagnoses to the nationwide Cancer Registry of Norway has been compulsory by national legislation. Patients are identified in the registry by their unique national registration number, assigned to all residents in Norway OPEN ACCESS and including date of birth. The cancer registry has maintained nearly 100% completeness for solid cancers, including breast cancer. 14 1 5

Prognosis in women with interval breast cancer: population based observational cohort study

Mette Kalager physician and postdoctoral researcher In 1996 the Norwegian breast cancer screening programme 25 Michael Bretthauer professor , Hans-Olov Adami started in four counties and then expanded gradually, county by

county, over the course of nine years. 4 1 3 Since 2005, all women 1 in Norwayof aged 50-69 years are invited to mammography Department Epidemiology,HarvardSchool of Public Health, Boston, USA; 3 4 Channing screening every two years. The Central Population Register ofNorway; From the screening database we further retrieved dataand on tumour Oslo, Norway; Department of Clinical Research, Telemark Hospital, Skien, Laboratory, Department of Medicine, Brigham 2 2 ); 5 Department of Organ NorwayHospital, identifies women eligible screening grade (based on the Nottingham grading system, I-III Womens Harvard Medical School, for Boston, USA; by their national Transplantation, Oslo University Hospital, Rikshospitalet, Oslo, 21 6 registration number. Invitations are posted to each eligible oestrogen and progesterone receptor status ; pTNM Norway; Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden woman, suggesting an appointment time. 1 6 Two radiologists classification; and dates of invitation and attendance to the independently read two view mammograms (craniocaudal views screening programme. We classified oestrogen and progesterone and mediolateral oblique views) in accordance with European receptors as positive (=10% positive staining) or negative (<10% Abstract Introduction 1 7 guidelines for quality assurance, which are classified according positive staining). These data were not available from the Objective To compare the prognosis in women with interval breast to a five point interpretation scale reflecting the probability of When mammography registrys database andscreening as a result programmes are not available are fully for women cancer (cancer detected after a normal screening mammogram and 18 cancer. The decision as to whether further diagnostic implemented, not invited to mammography interval cancersscreening. comprise asubstantial We defined screening proportion before the next scheduled mammogram) with breast cancer detected examinations are necessary is based on the consensus of two of incident rounds by county breast cancers. rather than Interval by individual cancers may women. have The been among women not yet invited to mammography screening experienced radiologists. After this final decision, no further overlooked follow-up period at the was last from mammography 1 January 1996 examination to 31 December or become (non-screened). diagnostic tests are done before the next scheduled screening apparent 2006. Linkage because to the they Central grew so Population rapidly that Register the detectable of Norway Design Population based observational study. invitation. preclinical and the National phase Death (sojourn Register time) was allowed shorter censoring than the atscreening date of interval. emigration or death. For the purpose of this study we defined Setting Norwegian breast cancer screening programme, implemented For the purpose of the study we classified women as having two study cohorts: in differentcancer counties if from 1996 cancer to 2005. was diagnosed within two years interval breast Because interval breast cancers in some studies on average are Interval group comprising women with a first 12 and two months the last screening mammogram but larger, of acancer more advanced stage, 1 andall express proliferative Participants 7116of women with normal a diagnosis of breast cancer at age 50 diagnosis of invasive breast cancer during 1 the 3 interval before invitation to the breast next cancer screening. Hence the cohort of markers more than screen detected tumours, it has been to 72 years; 1816 had interval and 5300 had a diagnosis between screening rounds breast or breast cancer women withbut interval included only those who were suggested thattwo prognosis of interval cancers is diagnosed poorer of breast cancer had not cancers yet been invited to screening. within two years and two months after 3 the date of last invited participants screening programme but had than that of screen detected breast cancers. However,the prognostic Main outcome measures in the Characteristics of the breast tumours, and normal examination the breast cancer screening non-screened detected breast cancer. For the presentCox analysis studies may be misleadingin when comparing interval breast survival of the women usingKaplan Meier curves andmultivariable programme between 1996 and 2006. we extended the screening interval by two months because we cancers with screen detected breast cancers because the screen
proportional hazard models. Results

retrieved date of diagnosis from two databases that were not Although interval cancers on average were slightly larger than completely in agreement about the date; the additional two the cancers in women not invited to screening, the histological type or months made it possible to include all women classified as status of axilliary lymph nodes did not differ noticeably between the two having interval cancers in the screening database. Among groups. Among interval cancers, there were no appreciable trends in women who reach the upper age limit of 70 years for invitation size, nodal status, grade,or hormone receptor positivity associated with to screening, we defined interval cancers as those diagnosed time since thelast normal mammogram as a marker ofgrowthrate. After within two years and two months after the last normal screening 10 years of follow-up, the survival rates were 79.1% (95% confidence examination.
interval 75.4% to 82.3%) among women with interval cancers and76.8%

We further divided the cancers into groups (sixgroup month (75.3% to 78.2%) among women in the non-screened cancer

intervals) according to the time between the0.84 date diagnosis (hazard ratio 0.98, 95% confidence interval to of 1.15; P=0.53). and the date of the last normal screening to explore Analyses stratified by time since last examination, normal mammogram, age at the hypothesis that more rapidly growing cancers arising shortly diagnosis, or screening round showed similar results. after a normal screening mammogram have a poorer Conclusion The prognosis of women with interval breast cancers was prognosis. 8 1 9 20 Owing to increased experience by the the same as that of women with breast cancers diagnosed without radiologists, screening sensitivity might increase with increasing mammography screening. screening rounds. Under this assumption the number of cancers overlooked at screening mammography should decline with time since start of the screening programme. Thus a growing Hazard ratios were calculated using Cox proportional hazard proportion of cancers detected between scheduled screenings models. We used likelihood ratio statistics to compare groups. should be true interval cancers. If true interval cancers were a adjusted for ageBoston, at diagnosis by Boston, four categories: 50-54, more aggressive breast of cancers, in theory the survival Correspondence to: Msubtypeof Kalager Department Epidemiology, Harvard School of Public Health,We 677 Huntington Avenue, MA 02115, USA mkalager@hsph.harvard.edu 55-59, 60-64, and 65-72 years. We further adjusted for time
No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

detected breast cancers are group affected comprising by length biasall sampling, Non-screened cancer women with lead time bias, and overdiagnosis bias. Therefore the valid a first diagnosis of invasive breast cancer who had not yet comparison group for assessment of prognosis in women with been invited to the breast cancer screening programme interval breast1996 cancers is2006. non-screen detected cancers among between and women not invited to mammography screening, which are unaffected by the biases that screening entails. Comparisons Statistical analyses with historical groups, as in many previous studies, may also 2 test survival We the Pearsonbecause to comparetheinterval cancer group leadused to confounding from breast cancer has 4 5 with the non-screened cancer group according to the improved over time. Only a few, small studies have compared characteristics the tumours, and we usedcancer a linear regression the survival of of women with interval breast with those model to test trends across time intervals from a last normal with non-screen detected breast cancer, with inconsistent 3 screening result. Using life table techniques we calculated breast findings. 6 -1 cancer specific survival survival of rates, In this population based rates studyand we overall took advantage the illustrated by Kaplan-Meier plots, and we compared these rates nationwide breast cancer screening programme in Norway, using the log rank test. Censoring occurred at date of emigration, which has been gradually implemented over a nine year period. 1 3 date of death from causes other than breast cancer, or the end This staggered roll out allowed a comparison between women of the follow-up period (31 December 2006), whichever came with interval breast cancer and those with breast cancer first.

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 3 of 10

RESEARCH

trends, county of residence, and time since last normal screening result. Because survival from breast cancer differed between counties, we adjusted for county and time trends by including county specific trend variables in the model. 4 We did not adjust for stage at diagnosis owing to the likelihood of stage migration, 23 but we carried out further analyses stratified by stage. For interval cancers only, we carried out secondary analyses to examine the association and possible interaction of time since last normal screening result and screening round. The proportional hazards assumption was tested by both graphical methods and Schoenfeld residuals and it was achieved. All test statistics were two tailed, and we considered P values <0.05 to be significant. Calculations were done with the statistical package Stata 10.0.

for age at diagnosis, year of diagnosis, and county of residency (data not shown). A secondary analysis restricted to interval cancers and stratified by screening round showed no evidence that survival was associated with screening round, either overall or after stratification by time since last normal screening result (table 4 ). Furthermore, after up to four screening rounds, the incidence of interval cancers was not associated with number of screening rounds. When the analysis was restricted to counties with a minimum of two years and two months of follow-up after examination, the incidence per 100 000 woman years was 163.4 in the first screening round, 162.5 in the second, 193.3 in the third, and 166.5 in the fourth.

Results
Table 1 summarises the characteristics of the 1816 women in the interval cancer group and the 5300 women in the non-screened cancer group. The mean age at diagnosis was similar between the groups, whereas the mean follow-up time was 3.6 (SD 2.6, maximum 10.6) years for the interval cancer group and 6.3 (SD 3.0, maximum 11.2) years for the non-screened cancer group. Compared with the non-screened cancer group, the interval cancer group had a slightly higher proportion of lobular cancers, largetumours (>20 mm diameter), negative axillary lymph nodes, and stage II rather than stage I disease (table 1). The proportion of women who had a sentinel node biopsy was about three times higher in the interval cancer group than in the non-screened cancer group. Adjuvant tamoxifen was given to 701 (38.6%) of the women in theinterval cancer group and 1912 (36.1%) in the non-screened cancer group. Table 2 shows the characteristics of interval breast cancers by six month intervals from date ofthelast normal screening result. The number of interval cancers increased with increasing time after a normal result. The mean tumour diameter for interval cancers increased by only 2.2 mm during the two years after a normal screening result (P for trend 0.03). There was no evidence that any other tumour characteristics varied by time since last normal screening result. The results were essentially the same with shorter intervals (data not shown). Cumulative breast cancer survival did not differ between the two groups (P=0.53, fig 1 ). After 10 years of follow-up, the survival rate among women with interval cancers was 79.1% (95% confidence interval 75.4% to 82.3%) and among women with cancers in the non-screened group was 76.8% (75.3% to 78.2%). Five year and 10 year age adjusted survival estimates by tumour stage at diagnosis were also similar for the study groups (data not shown). Cumulative overall survival did not differ between the two groups (P=0.67, fig 2 ). After 10 years of follow-up, the overall survival rate among women with interval cancers was 72.6% (68.5% to 76.3%) and among women with cancers in the non-screened group was 68.9% (67.3% to 70.5%). Table 3 shows data derived from Cox proportional hazards model analyses of age at diagnosis and time since last normal screening result as possible determinants of a difference in survival between the cancer groups. Breast cancer specific mortality did not differ between the interval cancer group and the non-screened cancer group (hazard ratio 0.98, 95% confidence interval 0.84 to 1.15, P=0.81). There was no association between time since last normal screening result and survival (table 3). The results did not change after adjustment

Discussion
The survival of women with adiagnosis of interval breast cancer after a normal mammogram is similar to that of women with breast cancer diagnosed before an invitation to a breast cancer screening programme. Contrary to our a priori hypothesis, we found no evidence that tumours that become clinically evident shortly after the last normal screening result were more aggressive in terms of larger size, higher grade, higher proportion of node metastases, or lower survival than non-screen detected breast cancers. Among the interval cancers, average tumour size increased slightly over time since last normal screening result, but no other characteristics of the tumour or risk of dying from breast cancer varied by time since the last normal screening result. Although several investigators have examined whether interval cancers are associated with poor survival, controversy remains. Previous studies have been limited by small sample size 7- 10 2 0 24 - 27 ; invalid comparison groups, notably screen detected cancers 9 25 - 27 ; or use of historical controls. 8 1 1 1 2 Furthermore, the findings in these studies were inconsistent. Analysis based on randomised trials of mammography screening found that survival with interval cancers was similar, 7 27 better, 20 or poorer 24 than survival with non-screen detected cancers. Although studies based on randomised trials have a valid comparison group, chance could explain the inconsistent findings because the sample sizes in all 7 2 0 24 2 8 the studies were limited (<100 interval cancers). In observational studies, where survival rates of women with interval cancers are compared with those of women with screen detected cancers or historical controls, or both, survival associated with interval cancers was similar to that of non-screen detected cancers in some studies, 9 1 1 2 6 worse in others, 1 4 and worse than screen detected cancers but better than clinically detected cancers in studies where interval cancers were compared with both screen detected and historical controls. 1 0 12 1 4 2 0 However, in these studies comparisons are likely to be confounded by lead time, overdiagnosis, and 4511 temporal trends in survival from breast cancer. The only remaining alternative is contemporary patients unaffected by mammography screening. Theoretically, the ideal comparison group would be breast cancer diagnosed among unscreened women who would have attended screening if they had been invited. Such a design would eliminate the potential confounding by factors that affect both attendance and outcome. However such a study is not feasible; to our knowledge there is limited reason in the Norwegian healthcare system to believe in any substantial bias through this mechanism, and this bias, if it exists, would affect only a small proportion of incident breast cancers because the attendance rate in the Norwegian mammography screening programme is high (77% among invited women). Overall survival was similar between women
Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 4 of 10

RESEARCH

with interval cancers and those with non-screen detected cancers (fig 2), indicating that the results were not influenced by the potential selection bias among women with interval cancer.

Funding: This study was fundedby researchgrants fromtheNorwegian Research Council and Frontier Science. The funders had no role in the design and conduct of the study; the collection,management, analysis, and interpretation of the data; and the preparation, review, or approval of the manuscript. Competing interests: All authors have completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work. Ethical approval: The researchprotocol was approvedby the Norwegian Social Science Data Services 2008. Individual informed consent was not requested. Data sharing: No additional data available.
1 Col let t K , St ef ans s on I M , E ide J, Br aat en A , Wang H, E ide GE , et a l. A ba sal ep ithe lial ph enot ype is mo re f r equent i n i nter v al br eas t c anc er s c om par ed wit h sc r een det ec ted t um or s . Canc er E pid emi ol Bi oma rk er s P r ev 200 5;14: 1108 -1 2. J Nat l Canc er I ns t 2 G illi and FD, J os t e N, S t auber PM , Hun t WC, Ros enb er g R, R edlic h G, et al. B iolo gic c har ac t er is ti cs of int er v al and s cr een -d etec t ed br eas t c anc er s. 20 00;9 2:743 -9 . 3 S iht o H, Lundi n J , Le htim k i T. M ol ecu lar s ubt y pes o f br eas t c anc er d etec t ed in m am mo gr aphy s c r eeni ng an d ou ts ide of s c r eening . Clin Canc er Res 20 08;14 :410 3- 10. 4 K alag er M, Hal dor se n T, B r ett hauer M , Hoff G , T hor es en S O, A dam i HO. I mpr ov ed br eas t c anc er su rv iv al f ollowi ng i ntr oduc t ion of an o rg aniz ed m am m ogr aphy s c r eeni ng pr ogr am am ong bot h s c r eened and uns c r eened wom en: a populat ion -b ase d c ohor t s t udy . Can ce r Res 2009; 1:R4 4. 5 Z ac k ri ss on S , J anz on L, M anjer J, And er ss on I . Im pr ov ed s ur vi va l r at e for wom en w ith i nter v al br eas t c anc er r es ults fr om the br eas t c anc er sc r een ing pr ogr am me in M alm , S wede n 197 6- 1999. J Med S c r een 2 007; 14:1 38- 43. 6 W ang H, B ju rs t am N, B j rn dal H, Br aat en A, E r ik s en L, S kaa ne P , et al . Int er v al c anc er s i n t he Nor wegian br eas t c anc er s cr een ing pr ogr am : fr equ enc y, ch ar act er is t ic s and us e of HRT. I nt J C anc er 200 1;94: 594- 8. 7 Hol mb er g LH, Ada mi HO , Tabar L, B er gs tr m R. S ur vi va l i n br ea st can ce r dia gnos ed be tween m am mog ra phic s c r eeni ng e xam in atio ns . i n t he DOM s c r eenin g pr og ra mm e. i nter m edia te out co me. J S ur g O nc ol E ur J Canc er 1996; 63:1 41- 4. Lanc et 198 6;2: 27- 30. 8 B r ek elm ans CT, P eeter s P H, Deur enber g J J , Col lett e HJ . Sur v iv al in int er va l br eas t canc er 1995; 31:1 830- 5. B re as t

Strengths and limitations of this study

To date, this population based study is the largest in the area of interval breast cancer and has the longest follow-up. The staggered introduction of the screening programme and the comparison of interval cancers with non-screen detected breast cancer avoided major biases. This study, however, has several potential limitations. Because only a few of the interval cancers had been individually reviewed, we were unable to distinguish true interval cancers from those overlooked when themammograms were examined. In a previous review in Norway comprising around 200 interval cancers, 35% were reinterpreted as overlooked, 23% showed minimal signs of malignancy, and 42% were true interval cancers, 2 9 30 similar to the results of other studies. 31 3 2 The overlooked tumours were on averagelargerand more often node positive than the true interval cancers. This could be because the affected women were reassured by the last normal mammogram result and therefore delayed seeking medical care. However, other studies have found no differences in survival 3334 between interval cancers classified as true or overlooked. In each Norwegian county, introduction of the breast cancer screening programme was preceded by the establishment of specialised multidisciplinary teams. 4 1 3 The goal of these teams was to provide thebest possible management of all women with newly diagnosed breast cancers in the county, regardless of whether the cancer was diagnosed by mammography screening. We have shown previously that this optimised management entailed a substantial reduction in mortality also among women with breast cancer not diagnosed by mammography screening. 4 1 3 Owing to the design of our current study, only women with interval cancers were managed by multidisciplinary teams. Hence it is conceivable that better treatment eliminated altogether a poorer prognosis among the women with interval cancers than among those with non-screen detected cancers. The possible influence of such confounding could not be tested in our study because it would have required more detailed individual data on prognostic factors, treatment, and overall management. However, the similar proportion of women in each group who received adjuvant tamoxifen treatments provides some reassurance against a major confounding.

9 S c hr en AA , Wobb es T, v an d er S luis RF . Int er v al c ar c ino mas o f t he br eas t: a g ro up wit h 10 Col lins S , Woo dman CB J, T hr elf all A, P r ior P . S ur vi va l r at es fr om i nter v al ca nc er s in NHS br ea st sc r eeni ng p ro gr amm e. B M J 199 8;316 :832 -3 . B re as t 11 B or ds P, J ons s on H, Ny st r m L, Lenn er P . S ur vi va l f r om inv as iv e br eas t ca nce r am ong i nter v al c as es in t he mam m ogr aphy sc r eeni ng pr ogr amm es of nor t her n S wede n. 20 07;1 6:47- 54 . 12 La wr enc e G , OS ul liv an E, Kea ri ns O , Tappe nden N, M ar tin K , Wal lis M . S c r eenin g hi st or ies of inv as iv e br eas t c anc er s d iagnos ed 1989- 20 06 i n t he Wes t M idla nds , UK : v ar ia tion wit h ti me and im pac t on 10- y ear s ur v iv al. J M ed S c r een 2009; 16: 186- 92. 13 K ala ger M , Z elen M, Langm ar k F , Ad ami HO . Ef f ec t of s c re ening m amm ogr aph y on br ea st - c anc er m or t alit y in Nor way . N E ngl J M ed 2010; 363: 1203- 10. 14 La rs en I K, Sm s tu en M , J ohan nes en T B , Langm ar k F , P ar ki n DM , B r ay F , et al . Data qu alit y at t he c anc er Reg is tr y o f Nor way : an ov er vi ew o f co mpa ra bilit y , c om plet enes s , v ali dit y and t im elines s . E ur J Canc er 2009; 45:1 218- 31. I nt J Canc er 15 T ingul st ad S , Haldor s en T, N or st ein J , Hagen B , S kj eldes t ad FE . Com plet enes s a nd ac c ur ac y o f r egis tr at ion of ov ar ia n c anc er in the c anc er r eg is tr y of Nor way . 20 02;9 8:907 -1 1. 16 W ang H, K r es en R, Her v ik A , Th or es en S . M amm ogr aph y s cr ee ning in N or way : re sul ts f r om t he fir s t s c r eening r ound in f our c ount ies and co st eff ec ti ve nes s of a m odeled na tio nwide s cr ee ning. Can ce r Caus es Contr ol 20 01;12 :39- 45 . 17 P er r y N, Br oed er s M , de Wolf C. Eur op ean g uideli nes f or q ualit y as s ur anc e in br eas t c anc er sc r een ing and d iagnos is . 2nd edn. O ff ic e for off ic ial publ ic ati ons of t he Eu ro pean Com m unit ies , 19 96. 18 E r z aas A . Q ualit y as s ur anc e man ual of th e nor we gian br eas t c anc er s c r eening pr ogr am [ Nor wegi an]. T he Canc er Re gis tr y of Nor way ; 2003 . www.k r ef tr egi s ter et .no . 19 Cow an WK , A ngus B , G ra y J C, Lunt LG, A l- T ami mi S R. A s tudy of int er va l br eas t c anc er wi thi n t he NHS br eas t s c r eenin g pr ogr am m e. a r andom is ed br eas t c anc er sc r eeni ng tr ial in S to ck hol m. 19 92;2 4:11- 6. 21 E dge SB , By r d DR, Com pt on CC, Fr it z AG , Gr eene F L, Tr ott i A , eds . AJ CC c anc er s t aging m anua l, 7t h ed n. S pr inger - V er lag, 20 10. 22 G alea MH, B la mey RW, E ls ton CE , El lis I O. T he No tt ingham P r ogno st ic I ndex in pr i mar y br ea st can ce r. Br eas t Ca nce r Res Tr eat 19 92;22 :207 -1 9. N 23 F eins t ein AR , So si n DM , We lls C K. T he Will Rog er s pheno men on. S tage m igr at ion and ne w di agnos ti c t ec hniques as a s our c e o f mi sl eading s t ati st ic s f or s ur v iv al in c anc er . E ngl J Med 1985 ;312: 1604 -8 . BMJ 24 A nder s s on I , A s pgr en K , Ja nzo n L, Landber g T , L indholm K , Linell F, et al. Ma mm ogr aphic s c r eenin g an d m or ta lit y f ro m br eas t c anc er : t he M alm m amm ogr aph ic t r ial. J C lin Pa thol 200 0;53 :140 -6 . B r eas t Ca nce r Res Tr eat

Conclusions
We conclude that tumours associated with interval breast cancers are more likely to be larger than those diagnosed in the absence of mammography screening; but they have strikingly similar survival outcomes. Furthermore, the characteristics of the tumours (except for an increase of 2.2 mm in diameter) and the prognosis of women with interval cancers were not associated with time since last mammography. These findings challenge the theory of a strong correlation between growth rate and metastatic behaviour. Our study provides no compelling support for moreaggressive primary treatment of interval breast cancers than non-screen detected cancers with similar prognostic features. 1 4
Contributors: MK designed the study and did the statistical analysis. MK had full access to the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. MK, RT, MB, and HOA interpreted the data and cowrote and edited the paper.

20 F r is ell J , v on Ros en A , Wie ge M , Nil ss on B , Gol dma n S . I nter v al c anc er s and s ur v iv al in

19 88;2 9:943 -8 . 25 V it ak B , S tl O, M ns on J C, T hom as B A, A r nes s on LG , Ek el und L, et al . Int er v al c anc er s an d c anc er s in non- at tend er s in t he s ter g tlan d m am mogr ap hic s c r eening pr ogr am me . Dur a tion bet ween s cr een ing and d iagnos is , s - phas e fr ac t ion and d is tant r ec ur r enc e. J Canc er 1 997;3 3:14 53- 60. 26 Ray s on D, P ay ne JI , A bdol ell M, Bar ne s P J, Mac I nto sh RF , Fol ey T , et al. Co mpa ri so n of cl inic al- pa thol ogic cha ra ct er is t ic s and out c ome s of t r ue i nter v al and s c r een- det ec ted E ur

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 5 of 10

RESEARCH

What is already known on this topic

Pre vi ou s r an do mi se d t ria ls o n ma mmo g rap h y s cre en in g f ou n d th a t in t erv al b re ast ca nce rs w ere a sso ci at e d w it h sim il ar , be t t er, or po o rer su rviv al com pa re d w it h no n -scre e ne d b re as t ca n ce rs T he se in co nsi st e nt f in di ng s ca n b e e xp la in e d b y sma ll sa mp le size s (< 1 00 in t erv al can ce rs) Obs erv at i on al st ud ie s su g ge st e d t ha t in te rv al c an ce rs w ere a sso cia t ed w it h p oo r su rvi va l b ut w ere li mit e d b y sm al l sa mp le si ze a nd in va li d co mp ar iso n g ro up s

What this study adds

I nt e rva l br ea st ca nce rs w e re m ore li ke ly t o b e l a rge r t ha n b re as t c an ce rs d ia gn os ed in th e a b se nce o f m am mo gra p hy s cre en in g Su rvi val ou t co me s be t we e n t h e t wo ca nc er g ro up s w ere , ho w eve r, st ri kin g ly si mi la r Our st u dy p ro vid e s n o co m pe ll in g su pp o rt f o r m or e ag g res siv e pr ima ry t re at m en t o f i nt e rva l bre a st c an ce rs t ha n n on -sc ree n in g ca nc ers wi t h si mi la r pro g no st ic fe a tu re s

i nv asi v e br ea st c an cer am ong par ti ci pant s of a Canadi an br eas t s cr ee ning pr ogr am : a ne st ed c as e- co ntr ol s tu dy. Clin B re as t Canc er 20 11;1 1:27- 32 . S ur ger y 19 83;9 4:54 3- 7. 27 De Gr oot e R, Rus h B F J r , Mi laz z o J , W ar den MJ , Roc k o J M. I nt er va l br ea st can ce r: a m or e aggr es s iv e sub se t of br eas t neop las ias . on br eas t c anc er m or tali ty . J Natl Canc er In st

f indi ngs at the lat es t s c r eenin g: a c ompa ri s on of tr ue int er va l and mis s ed inte rv al c anc er s . E ur Radiol 199 9;9: 460- 9. J Med S c r een 34 P or ter G J R, E v ans A J, B ur rel l A J, Lee AHS , El lis IO , Ch akr abar t i J . Int er val br eas t c anc er s : pr og nos tic fe atur es and su rv i val by s ubty pe and t ime s inc e las t s c r een. 20 06;1 3:115 -2 2.

28 S hap ir o S , Ven et W , St r ax P , Ven et L , Ro es er R. Ten- t o f our te en- ye ar ef fec t of s cr ee ning 1 982; 69:34 9- 55. 29 Ho fv ind S , Sk aa ne P , V it ak B , Wan g H, T hor es en S, E r ik s en L, et al. I nf luenc e of r ev iew de si gn on pe rc en tages of m is s ed i nte rv al br eas t c anc er s : a r etr os pe ct s t udy o f int er va l c anc er s in a po pulat ion- bas ed s c re ening pr ogr am . of int er va l br ea st can ce rs . A c ta Radi ol Radi ology 2008; 49: 975- 81. 2005; 237: 437- 43. 30 Ho fv ind S , Gel ler B , S k aane P . M amm ogr aph ic f eat ur es and his t opat hologi c al f ind ings 31 V it ak B . I nt er va l c anc er s in the Os t er gtl and Mam m ogr aphi c S c re ening P r ogr am: r adi ologi ca l a naly s is . E ur Radi ol 1998; 8:6 39- 46. 32 Do min go L , Sa la M, S er v it ja S, Cor omi nas J M , Fer r er F, M ar t nez J , et al. P hen oty pi c c har ac t er iz at ions and r is k fa ct or s f or int er v al br eas t c anc er s in a popu lat ion- bas ed br eas t c anc er s c r eening pr ogr am in B ar c elona, S pain. Canc er Caus es Con tr ol 2 010;2 1:1155 -6 4. 33 V it ak B , Ol se n K E , M nso n J C, A r nes s on LG , St l O. T umo r c har ac ter i st ic s an d s ur v iv al i n pa tie nts with in vas i ve int er v al br eas t c anc er c la ss if ied ac c or ding t o m am mo gr aphic

Accepted: Cite this as:

1 November 2012 BMJ 2012;345:e7536

T his is an o pen - access ar t icl e di str ib ut ed u nde r t he t er m s of th e Cr ea ti ve Com m o ns At tr i but io n Non - com m er ci al Li cense , whi ch per m i ts use , dist r ib ut ion , an d r epr o duct io n i n an y m e diu m , pr o vide d th e or ig ina l wor k is pr op er ly cit ed , t he use i s n on com m er ci al an d is ot her w is e in c om pl ian c e wi th t he lic e ns e. S e e: ht t p: // c r eat iv e c om m on s .o r g/ lic e ns es / by nc/ 2. 0/ and h tt p :/ / cr eat ive com m on s.o r g/ lice nses/ by- n c/2 .0 /l ega lcod e.

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 6 of 10

RESEARCH

Tables
Table 1 | Characteristics of women with breast cancer in interval cancer group and in non-screened cancer group. Values are numbers (percentages) unless stated otherwise
C ha r ac te r i st ic s M e an (S D ) ag e a t d i ag n o sis (ye a rs) H i st o lo g ica l ty p e: D u ct a l L ob u la r Ot h e r Tu m ou r s iz e* : 1 (< 2 0 m m) 2 (> 2 0- 50 m m) 3 (> 5 0 m m) 4 (in g ro w t h) U n kn ow n N o d al st a t us *: Po s it iv e N e ga t iv e M iss in g S t ag e : I II III IV S en t i ne l n o d e b i op sy Ta m ox if e n ad m in is t er ed To t a l N o of d e at h s D e a th s f r om b re as t c a nc er 7 5 0 ( 41 . 5 ) 9 1 3 ( 50 . 6 ) 5 6 ( 3. 1 ) 8 7 ( 4. 8 ) 8 6 9 ( 47 . 9 ) 7 0 1 ( 38 . 6 ) 2 4 6 ( 13 . 5 ) 1 9 4 ( 10 . 7 )
21

I nt er v a l c a nc e r gr oup (n= 1 8 1 6) 5 9 .4 (5 . 7 )

N on-s c re e ne d c anc e r g roup ( n= 5 30 0 ) 6 0 .1 (7 . 1 )

P v a lu e

1 3 7 7 (7 5 . 8 ) 2 4 5 ( 13 . 5 ) 1 9 4 ( 10 . 7 )

4 1 76 (7 8 . 8) 5 3 1 (1 0 . 0 ) 5 9 3 (1 1 . 2 )

<0 . 0 0 1

9 6 2 ( 56 . 1 ) 5 7 4 ( 33 . 5 ) 1 3 7 (8 . 0 ) 4 2 ( 2. 5 ) 1 0 1 ( 5. 6 )

2 4 64 (5 9 . 9) 1 3 44 (3 2 . 7) 1 06 (2 . 6 ) 1 97 (4 . 8 ) 1 1 8 9 (2 2 . 4 )

<0 . 0 0 1

9 1 0 ( 54 . 4 ) 7 6 4 ( 45 . 6 ) 1 4 2 ( 7. 8 )

2 8 05 (5 8 . 7) 2 4 91 (4 1 . 3) 5 2 0 (9 . 8 )

<0 . 0 0 1

2 6 07 (4 9 . 4) 2 2 36 (4 2 . 4) 1 53 (2 . 9 ) 2 79 (5 . 3 ) 7 9 8 (1 5 . 0 ) 1 9 12 (3 6 . 1) 1 3 07 (2 4 . 7) 9 7 9 (1 8 . 5 )

<0 . 0 0 1

< 0 . 00 1 < 0 . 00 1 < 0 . 00 1 < 0 . 00 1

* Ba se d o n p a t ho l og i ca l t u mo u r, n od e , an d me t a st a si s cl a ssi f ic at i on . % o f t o t a l b re as t ca nc e rs.

I , l o ca li se d br ea st ca n ce r; I I , l ym p h n o de p os it i ve ; I I I , gr ow t h in t o s ki n o r ch e st w al l; a n d I V , m e ta s ta s is. M is sin g in f o rma t i on o n s ta g e : 1 0 in t e rva l c an ce rs an d 25 n o n- scr ee n e d c an c er s.

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 7 of 10

RESEARCH

Table 2 | Tumour characteristics of interval breast cancers in six month intervals from last normal screening result to diagnoses. Values are numbers (percentages) unless stated otherwise
Ti me si nc e l as t no rma l s cr e en in g re s ul t (mont hs) C ha r ac te r i st ic s 19 -2 4 (n= 5 8 4) 0- 6 (n = 18 1 ) M e an (S D ) t um o ur siz e ( mm ) N o d e n e ga t i ve M is si ng n od e st a t us * Gr ad e I Gr ad e I I Gr ad e I I I M is si ng g ra de * Oe s tr og e n re ce pt o r s ta t u s: Po s it iv e N e ga t iv e M iss in g da t a * P ro ge s te ro n e re ce pt o r s ta t u s: Po s it iv e N e ga t iv e M iss in g da t a * * % of t o t al b re as t c a nc er s. = 1 0% p o si t ive st a i ni n g. < 1 0% p o si t ive st a i ni n g.
21

7 -1 2 (n= 4 9 7) 1 9. 8 (1 1 .8 ) 25 3 (5 4 .5 ) 3 3 (6 . 6) 81 (1 9 . 6) 20 6 (4 9 .9 ) 12 6 (3 0 .5 ) 84 (1 6 . 9)

13 -1 8 (n= 5 5 4) 2 0. 6 (1 2 . 5) 2 67 (5 2 . 8) 48 (8 . 7 ) 92 (2 0 . 2) 2 14 (4 6 . 9) 1 50 (3 2 . 9) 98 (1 7 . 7) 2 1. 5 (1 3 .4 ) 3 04 (5 6 . 4) 45 (7 . 7 ) 1 30 (2 7 . 0) 2 13 (4 4 . 3) 1 38 (2 8 . 7) 1 03 (1 7 . 6)

1 9. 3 (1 2 . 2) 8 6 (5 2 . 1 ) 16 (8 . 8 ) 3 5 (2 4 . 0 ) 6 7 (4 5 . 9 ) 4 4 (3 0 . 1 ) 3 5 (1 9 . 3 )

9 2 (7 4 . 8 ) 3 1 (2 5 . 5 ) 5 8 (3 2 . 0 )

22 4 (6 9 .6 ) 98 (3 0 . 4) 17 5 (3 5 .2 )

2 51 (7 1 . 7) 99 (2 8 . 3) 2 04 (3 6 . 8)

2 45 (7 3 . 8) 87 (2 6 . 2) 2 52 (4 3 . 2)

6 4 (5 2 . 9 ) 5 7 (4 7 . 1 ) 6 0 (3 3 . 2 )

16 2 (5 1 .1 ) 15 5 (4 8 .9 ) 18 0 (3 6 .2 )

1 81 (5 3 . 2) 1 59 (4 6 . 8) 2 14 (3 8 . 6)

1 74 (5 3 . 2) 1 53 (4 6 . 8) 2 57 (4 4 . 0)

21

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 8 of 10

RESEARCH

Table 3 | Hazard ratios (95% confidence intervals) for breast cancer specific mortality between breast cancer groups overall and across age categories
A ge a t di ag nos is (y e a rs ) Ov er a l l H az a rd r a ti o* V ar i a bl es N o n -sc re en e d c an ce r g ro u p ( 95 % C I ) 1 (re f e re n ce ) (n = 5 3 00 ca n ce rs) I n t er va l c an ce r g ro u p 0. 9 8 (0 . 84 t o 1 . 15 ) (n = 1 8 16 ca n ce rs) M o nt h s s in ce l as t scr ee n : 0 -6 1. 0 9 (0 . 70 t o 1 . 67 ) 7 -1 2 0. 9 7 (0 . 74 t o 1 . 28 ) 1 3- 18 1. 0 6 (0 . 82 t o 1 . 36 ) 1 9- 26 0. 8 7 (0 . 66 t o 1 . 15 ) * Ag e ad j us t ed . C a te g o ris e d a s t i me si nc e l a st n or ma l m a mm o gr am (t i me si nc e l a st sc re en ) i n s ix mo n t h i nt e rv al s. P f o r t re n d. 0. 5 3 0 . 6 8 (0 .2 8 t o 1 .6 5 ) 0 . 8 1 (0 .4 3 t o 1 .5 3 ) 1 . 0 0 (0 .5 8 t o 1 .7 6 ) 0 . 6 6 (0 .3 1 t o 1 .4 0 ) 0. 6 1 1 . 61 (0 . 8 0 t o 3 . 2 9) 1 . 19 (0 . 7 7 t o 1 . 8 5) 1 . 04 (0 . 6 7 t o 1 . 6 4) 1 . 04 (0 . 6 6 t o 1 . 6 5) 0 . 8 4 1 . 1 9 ( 0. 4 4 to 3. 2 1 ) 0 . 8 2 ( 0. 4 2 to 1. 6 0 ) 1 . 3 7 ( 0. 7 2 to 2. 0 9 ) 0 . 5 2 ( 0. 2 3 to 1. 1 9 ) 0.78 1. 1 2 (0 . 42 t o 3 . 00 ) 0. 9 7 (0 . 56 t o 1 . 69 ) 1. 0 1 (0 . 62 t o 1 . 68 ) 1. 0 4 (0 . 66 t o 1 . 63 ) 0 . 91 0 . 81 P v al ue 5 0- 54 H a za rd r a ti o (9 5 % C I ) 1 (re f e re nc e ) (n = 1 50 7 ca nc e rs) 0 . 8 1 (0 .5 7 t o 1 .1 6 ) (n = 4 2 9 ca n ce rs) 0 . 25 P v a lu e 5 5 -5 9 H a za r d ra ti o (9 5% C I ) 1 (r ef e re n ce ) ( n= 1 5 5 8 ca n ce rs ) 1 . 13 (0 . 8 6 t o 1 . 4 9) (n = 57 8 c an c er s) 0. 3 7 0 . 8 9 ( 0. 6 1 to 1. 3 0 ) (n = 3 99 ca nc e rs) 0 .5 5 P v a l ue 60 -6 4 H az a rd r a ti o (9 5 % CI ) 1 (re f er en c e) (n = 9 63 ca nc e rs) P va l ue 6 5 -7 2 H a za rd r a ti o ( 95 % C I ) 1 (re f e re n ce ) (n = 1 6 72 ca n ce rs) 1. 0 2 (0 . 76 t o 1 . 37 ) ( n= 4 1 0 ca n ce rs ) 0.89 P v a l ue

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 9 of 10

RESEARCH

Table 4 | Age adjusted hazard ratios (95% confidence intervals) for breast cancer specific mortality among women with interval breast cancers comparing screening round according to time since last normal screening result
Ti me s i nc e l a s t nor mal s c re e ni ng r es ul t (mo nths ) S cr e e ni ng rou nd (N o of c a nce r s ) 1 (n = 5 64 ) 2 (n = 4 75 ) 3 (n = 3 55 ) 4 (n = 2 20 ) 5 (n = 1 82 ) O ve r a ll 1 (re f e re nc e) 0 .8 9 (0 . 63 t o 1. 2 7 ) 0 .8 6 (0 . 58 t o 1. 2 7 ) 0 .6 9 (0 . 39 t o 1. 2 1 ) 0 .9 9 (0 . 42 t o 2. 3 1 ) 0- 6 1 (re f e re n ce ) 0 . 85 (0 . 2 8 t o 2 6 4 ) 1 . 07 (0 . 3 6 t o 3 2 3 ) 0 . 32 (0 . 0 4 t o 2 6 1 ) 1 . 19 (0 . 3 7 t o 9 8 8 ) 7 -1 2 1 (r ef e re n ce ) 1 . 0 6 (0 . 5 5 t o 2 . 06 ) 1 . 1 8 (0 . 5 8 t o 2 . 41 ) 0 . 6 6 (0 . 2 2 t o 1 . 94 ) 0 . 3 8 (0 . 0 5 t o 2 . 93 ) 1 3 -1 8 1 (r ef e re n ce ) 0 . 5 5 ( 0. 2 9 to 1 . 02 ) 0 . 3 9 ( 0. 1 7 to 0 . 89 ) 0 . 4 3 ( 0. 1 5 to 1 . 23 ) 1 . 3 9 ( 0. 4 0 to 4 . 80 ) 1 9 -2 4 1 ( re f er en ce ) 1 . 4 8 (0 .7 6 t o 2 . 8 7) 1 . 1 8 (0 .5 4 t o 2 . 6 0) 1 . 6 5 (0 .6 4 t o 4 . 2 9) P for tr e nd 0. 4 1 0. 7 2 0. 8 5 0. 7 4 0. 1 7

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe

BMJ 2012;345:e7536 doi: 10.1136/bmj.e7536 (Published 16 November 2012) Page 10 of 10

RESEARCH

Figures

Fig 1 Cumulative breast cancer survival plot for women with breast cancer by group

Fig 2 Cumulative overall survival plot for women with breast cancer by group

No com m e r cial r e use: Se e r ig ht s a nd r ep r int s ht tp :/ / www. bm j .co m /p er m issi ons

Su bscr ib e: h tt p: / /w ww.b m j. com / subscr ibe