Version of Record: https://www.sciencedirect.

com/science/article/pii/S0301211519301587
Manuscript_a51474c38e04a8652eafdb7d2e3c3283

Full title

Accuracy of clinical breast examination’s abnormalities for breast cancer screening: cross-

sectional study

Authors

Alexandre MALMARTEL 1, Arthur TRON 1, Ségolène CAULLIEZ 1

1
Department of General Practice, Faculté de medicine, Université Paris Descartes, Paris,

France

Corresponding author

Dr Alexandre MALMARTEL

Université de Médecine Paris Descartes – Site Cochin, Département de médecine Générale

24, rue du Faubourg Saint-Jacques

75014 Paris - France

e-mail : [email protected]

phone: +33 6 81 81 17 20

fax: +33 1 44 41 23 64

Funding :

The study did not receive funding.

Competing of interest:

The authors declared they have no competing interests.

1
© 2019 published by Elsevier. This manuscript is made available under the Elsevier user license
https://www.elsevier.com/open-access/userlicense/1.0/
Accuracy of clinical breast examination’s abnormalities for breast cancer screening:

cross-sectional study

Alexandre MALMARTEL 1, Arthur TRON 1, Ségolène CAULLIEZ 1

Abstract:

Objectives: The guidelines for breast cancer screening with clinical breast examination

(CBE) are diverging: CBE is recommended in France, whereas it is not recommended in the

United States and Canada, given the lack of clear benefit and the risk of overmedication. To

assess the accuracy of abnormalities found during CBE for in breast cancer screening.

Study design: A cross-sectional study included women over 18 years with no history of

breast cancer coming for a mammography at 3 ambulatory radiology practices in Paris. A

questionnaire collected the risk of breast cancer on mammography according to the Breast

Imaging-Reporting And Data System (Bi-RADS) (high risk: Bi-RADS 4 or 5 versus lower

risk: other Bi-RADS categories), the risk factors for breast cancer and the breast clinical

abnormalities (none, mass, skin abnormality, oedema, pain, nipple discharge, lymph nodes...)

For each abnormality, sensitivity, specificity, positive (PPV) and negative (NPV) predictive

values were calculated.

Result: Among the 3218 included patients (mean age: 55.1 +/-10 years), 713 (22.2%) had an

abnormal CBE and 133 (4.1%) had high-risk mammography. The sensitivity of CBE was

36%[28%;45%] and the specificity was 78%[77%;80%]. The PPV and NPV for each clinical

abnormality were low, except for nipple discharge, retraction and lymph nodes, for which the

PPV were 10.5[3.7;29.9], 6.6[1.4;31.6], and 5.0[1.5;17.1], respectively, but these

abnormalities were rare (0.5%, 0.2% and 0.5% respectively). These values were similar across

all age groups.

2
Conclusion: The accuracy of CBE for breast cancer screening appeared to be low which did

not support recommending regular CBE in France.

Key points:

• In this study, 4.1% women had a high-risk mammography (Bi-RADS 4 or 5)

• The sensitivity of clinical breast examination was 36% and its specificity was 78%

• Clinical abnormalities did not have sufficient accuracy to suspect breast cancer

• Nipple discharge had a high predictive positive value but was rarely present.

Keywords : breast cancer, clinical breast examination, screening, sensitivity and specificity

3
Introduction

Breast cancer was the most common cancer among women and the leading cause of cancer

mortality among women in France in 2012 [1]. The management of risk factors and the

implementation of prevention and screening actions are essential to prevent the progression of

this pathology and reduce breath cancer mortality [2,3]. Thus, the French Health Authority

(Haute Autorité de Santé, HAS) recommended an annual clinical breast examination (CBE)

performed by a general practitioner or a gynecologist for women from 25 years, possibly

combined with mammography follow-up according to breast cancer risk factors as part of

individual screening. For the individual screening, patients were referred by a physician,

mostly general practitioner or gynaecologist, for a mammography because of an abnormal

CBE or because of an estimated higher risk of breast cancer regardless the result of the CBE.

In addition, from the age of 50, patients were invited every 2 years to undergo a

mammography screening by an organized screening. They automatically receive a

mammography prescription from the French healthcare system. With this national program,

7,5% cancer were diagnosed in 1000 screened women in 2011-2012 [4] These

recommendations for screening and follow-up with annual CBE have been in place for many

years now based on the fact that breast cancer may be asymptomatic or manifested by the

presence of one or more clinical abnormalities [5]. Thus, French women 50 and 74 years can

undergo mammography with the organized screening or with the individual screening instead,

if they request it to their doctor even with a normal CBE.

However, Canadian and American guidelines no longer recommend systematic CBE,

although any symptomatic patient should be assessed promptly by a mammography. [6,7]. In

fact, some studies found that there was no clear benefit of regular CBE and that it could lead

to over-diagnosis, over-treatment and iatrogenic complications [8,9].

4
Thus, the clinical relevance of systematic CBE is little studied and its efficacy has not been

demonstrated. We hypothesis that CBE might not be reliable enough to be recommended for

breast cancer screening in France. The objective of our study was to assess the diagnostic

accuracy of the clinical abnormalities found during CBE, performed by a physician, as part of

the breast cancer screening.

Materials and methods

A multicenter, cross-sectional study included consecutively women over 18 years who came

to perform mammography in 3 radiology practices in Paris between July 2017 and January

2018. The mammography could be indicated because the patients were involved in the

organized screening, or in an individual screening (replacing the organized screening or

because of a higher individual risk of breast cancer) or because of an abnormal CBE. Patients

were informed of the study using an information sheet and consent was obtained prior to

mammography. Patients with a personal history of breast cancer were excluded from our

study. If necessary, the radiologists could, according to their current practice, complete the

mammography with an ultrasonography.

A medical questionnaire completed by the radiologist collected the age of the patient, the

screening method (individual or organized screening), the abnormalities found with the CBE

just before the mammography (none, breast mass, skin retraction, oedema, erythema, orange

peel, ulceration, pain, nipple discharge, lymph nodes, other), personal and family history of

breast cancer, number of pregnancies, and the risk of breast cancer assessed according to

American College of Radiology criteria (Bi-RADS) after the mammography [10,11]. In this

study, 15 radiologists included patients and conducted clinical examination just before the

mammography. They were specialised in senology and qualified to perform CBE.

5
The primary outcome was the Bi-RADS score for each mammography. This risk of breast

cancer was reclassified as high risk of malignancy for Bi-RADS 4 and 5, and lower risk for

other Bi-RADS categories (Bi-RADS 0, 1, 2 and 3).

As the prevalence of abnormal mammography was 12%[12] and the CBE sensitivity was

estimated to be near 60%[13,14], we needed to include 3100 patients to reach an absolute

precision of 0.05 with α equal to 5% according to Buderer’s formula[15].

Univariate analyses between breast cancer risk and explanatory variables were performed

with Chi-2 tests or Fisher’s test for categorical variables, and Student tests for continuous

variables.

The sensitivity (Se), specificity (Sp), positive predictive value (PPV) and negative predictive

value (NPV), as well as the positive (LR+) and negative (LR-) likelihood ratios for each of

the clinical abnormalities as a function of malignancy risk were calculated for all patients and

according to patient age. We considered that LR+ and LR- significantly contribute to the

diagnosis when they were over 5 and below 0.1, respectively. Statistical analyses were

performed using the R software (http://cran.r-project.org/).

Patients could refuse to participate or withdraw their consent at any time. The study protocol

was approved by the Institutional Review Board of CPP SUD EST 1 (IRB No 2018-24 Paris

Descartes RIPH 3). The study did not receive any funding.

Result

During the inclusion period, 3218 women were included (Figure 1). Their median age was 55

years (Interval Interquartile (IQR)= [48-63]), 957 patients (29.7%) were under 50 years of

6
age, 2190 (68.1%) were between 50 and 74 years, and 71 (2.2%) were over 74 years of age

(Table 1).

In our patients, 713 (22.2%) had an abnormal clinical examination. The main abnormalities

found in women with abnormal CBE were breast masses in 469 patients (14.6%), breast pain

in 126 patients (3.9%), nipple discharge in 16 women (0.5%), axillary nodes in 17 patients

(0.5%) (Figure 1).

According to table 1, 133 women (4.1%) had a high-risk mammography for breast

malignancy. Among these women with high-risk mammography, 48 (36.1%) had abnormal

CBE, compared to 665 (21.6%) among those with low to moderate risk mammography (p

<0.01). Associated factors with a high-risk mammography in multivariate analysis were the

presence a mass (OR=1.8; 95% Confidence Interval (95%CI)=[1.1;2.7]; p=0.01), a nipple

discharge (OR=11.3; 95%CI=[3.5;32.0]; p<0.01), a lymph node (OR=6.3; 95%CI=[1.7;20.2];

p<0.01), and nulliparous status (OR= 1.5; 95%CI=[ 1.0;2.1] ; p=0.04).

Patients performing mammography with the individual screening had more abnormal CBE

compared to patients in organized screening (24.3% versus 10.7 % ; p< 0.01) but they did not

have a more high-risk mammography (4.2% versus 3.5% ; p=0.52).

In the overall population, CBE sensitivity was 36% (95%CI=[28%; 45%]) and its specificity

was 78% (95%CI=[77%; 80%]). LR+ and LR- were low 1.7 (95%CI= [1.3; 2.1]) and 0.8

(95%CI=[0.7; 0.9]), respectively. Concerning the presence of breast mass, the results were

similar: the sensitivity was 26% (95%CI=[18%; 34%]), the specificity 86% (95%CI= [85%;

87%]). Sensitivity and specificity for nipple discharge were 4% (95%CI=[1%; 9%]) and

100% (95%CI=[99%; 100%]), for retraction 2% (95%CI=[0%; 5%]) and

100%(95%CI=[100%; 100%]), and for axillary nodes 2% (95%CI=[0%; 6%]) and 100%

(95%CI=[99%; 100%]), respectively. Their LR+ were 10.5 (95%CI=[3.7; 29.9]) , 6.6

7
(95%CI=[1.4; 31.6]) and 4.97 (95%CI=[1.5; 17.1)], respectively but these clinical

abnormalities were very rare (Table 2).

Among women under 50 years, CBE sensitivities were low, ranging from 2 to 38%, while

specificities were good, between 73% and 99%. However, likelihood ratios were, between 0.4

and 8.1 for LR+ and between 0.9 and 1.0 for LR- (Table 3).

In patients aged from 50 to 74 years, the overall results were similar. Sensitivities ranged from

1% to 36%, and specificities were between 80% and 100%. LR+ were high for nipple

discharge and retraction (LR+ =17.4 ; 95%CI=[2.9; 102.5] for both) (Table 4).

For patients the 71 patients aged over 74 years, the results were similar for any CBE

abnormalities combined: the sensitivity was 29% [4%; 19%], the specificity 88% [77%;

94%], the PPV 20% [3%; 56%] and the NPV 92% [82%; 97%], the LR+ 2.3 (95%CI=[0.6;

8.7]) and the LR- 0.8 (95%CI= 0.5; 1.3]). The small numbers of patients over 75 years did

not allow analysing each clinical abnormality individually.

Discussion

Summary

In this study, 22% of all patients had an abnormal clinical examination and 11% among

patients involved in the organized screening. CBE abnormalities had a good specificity

(between 78% and 100%) and a poor sensitivity (36% to 0%) in the included population to

detect high risk mammography, which was an original and relevant outcome for clinicians.

Indeed, for any anomaly combined, the sensitivity of the CBE was 36% and the specificity

78%. For women from 50 to 74 years, and those under 50 years, the sensitivity and the

specificity of the CBE were similar to that found for the overall population.

A breast mass found during the CBE did not present good accuracy. Whatever the age of the

women, its sensitivity was about 26%, its specificity 86% and the likelihood ratios were low

8
(LR+ = 1.8 and LR- = 0.8). In the overall study population, the LR+ and LR- of the CBE and

of each abnormality contributed little to the diagnosis, with the exception of nipple discharge,

retraction and axillary nodes, for which the LR+ were greater than 5. However, the frequency

of these 3 abnormalities seemed too low for breast cancer screening to rely on their presence.

Strengths and limitations

Our study was one of the rare recent studies evaluating the accuracy of the CBE by a

healthcare professional, and the only one conducted in French population although some

powerful studies already found low performance of CBE in other settings. One of the main

strengths of our study was the consecutive recruitment of patients limiting selection bias. CBE

is a difficult and clinically dependent examination. In order to overcome this disadvantage, we

deliberately limited the number of inclusion centers in order to obtain more reproducible

results. Thus, the CBE was performed by experts in senology from radiology practices, but it

is likely that CBE would be less effective if performed by a general practitioner.

However, our study has some weaknesses. First, we used the high-risk of malignancy

depending on the Bi-RADS classification on mammography as the primary endpoint, rather

than cancer diagnosis, because it was impossible to follow the patients to know the results of

any biopsies performed if they were indicated. Thus we could have overestimated high-risk

patients compared to the actual number of patients with breast cancer, increase the sensitivity

of CBE and lower its specificity. However, our criterion of judgment is clinically relevant

because the results of the mammography lead to a modification of the management by the

physician: in case of mammography with Bi-RADS 4 or 5, the physician will prescribe a

biopsy. Second, mammography may not be the reference for young women because of the

high breast density at these ages. Thus, according to their current practices, radiologists could

9
complete the mammography with a breast ultrasound before assessing the risk of malignancy.

Third, the radiology practices in our study were specialized senology centers. There may

therefore have been a selection bias within our study, with a female population that might be

more symptomatic than the general population. Similarly, the prevalence of high-risk

mammography was probably different from that general population. Given that many patients

were participating in individualized screening, the number of symptomatic patients with high-

risk mammography may also be partly due to missed screening opportunities with organized

screening in previous years. Thus, PPV and NPV values may be difficult to generalize.

However, the inclusion of patients in several centers may have reduced this bias. Fourth, the

low number of patients with skin abnormalities, erythema, ulceration, oedema, nipple

discharge, or skin retraction did not allow calculating every parameter in all age’s categories.

Finally, in women participating to an individual screening, it was not possible to differentiate

between women who were referred following the discovery of a clinical abnormality by a

professional and those referred for routine mammography as part of an individual screening,

but this data was not necessary to assess the diagnostic accuracy of CBE.

Comparison with existing literature

Despite only few studies have been conducted to evaluate the effectiveness of CBE alone, our

results are in line with the North-American guidelines and with the other studies. A British

study carried out in 1992, found a CBE sensitivity of 64% [13]. In the same year, a Canadian

study found a sensitivity of 69% in women aged 40-49 and 63% in women aged 50-59

[14].These higher sensitivities compared to the results of our study can be explained by the

time at which they were conducted. Radiological tools were less effective than today [16,17],

more cancers might be diagnosed following clinical examination, thus increasing the

sensitivity of the CBE. A study conducted in 2002 estimated a sensitivity of 35% [18] which

10
is consistent with our study and reinforces its validity. In primary care setting, our results are

consistent other studies finding that breast cancer was associated with breast lump and nipple

discharge[19,20]. Nevertheless, we did not find any increase of PPV of breast mass between

younger patients and patients over 70 years, probably because breast cancer effecting elderly

women in 2000-2009 were not discovered with breast screening programs at a younger age.

Until then, CBE was also offered to women aged 50 to 74 at the annual screening visit, to

minimize the risk of not detecting radio-occult cancers and because it could help to diagnose

more aggressive breast cancer than mammography alone [21]. But when the mammography’s

result was initially considered normal, an abnormality was detected at CBE in only 0.2% of

cases in 2010 [22]. However, CBE could remain relevant and should be performed in low and

middle income countries (LMICs) where the access to mammography may be more difficult.

In these countries, CBE can reduce delay to diagnose breast cancer and it is associated with

earlier stage cancer [23]. The average size of breast tumours at the time of their discovery in

France was 1.9 cm, whereas it was 5 cm in Tunisia for example [24]. Nevertheless,

mammography should be encouraged as biennial mammography reduced in breast cancer

deaths compared to annual CBE alone [25]. In the same way, self breast examination was not

recommended by North-American guidelines because it did not reduce breast cancer mortality

in clinical trials [6,7,26], but it could help to detect breast cancers at early stages in LMICs

[27].

Implications for research and practice

In the current state of knowledge, it is not possible to distinguish cancers that will evolve from

those that will not or only slightly evolve: all the lesions detected are therefore treated,

implying over-diagnosis and over-treatment. Various studies and meta-analyses estimate that

11
30% to 50% of diagnosed breast cancers are over-diagnosed [8,28]. In the United Kingdom,

the concept of "breast awareness", teaching patients “red flags” that the physicians can also

check, seems to have relevant results in terms of breast cancer screening [29]. Thus, it seems

essential to develop primary prevention measures for this cancer. In fact, the breast cancer

prevention strategy implemented in 2017 by the French National Institute for Cancer (INCa)

suggested two information and prevention medical consultations at 25 and 50 years [30].

They aimed at raising awareness of the risk factors for breast cancer, and means of prevention

such as food hygiene and regular physical activity [2,31]. These measures could help to

prevent 20,000 breast and colon cancers each year [32]. In patients between 50 and 74 years,

health educational programmes could also increase women’s breast cancer awareness and

influence their participation in screening [33,34].

Conclusion

CBE does not seem to provide reliable and reproducible additional information compared to

mammography. Our results were consistent with the other studies and were not in favour of

its application in France for the regular gynaecological follow-up of women between 50 and

74 years. CBE could have a better place in the follow-up of breast lesions but not for

screening or diagnosis. Nevertheless, the implementation of prevention measures and the

development individualized screening on each patients risk with screening tools [35] or

genomics [36] could reduce the incidence of this cancer and optimize screening by limiting

over-diagnosis.

12
Acknowledgement:

The authors wish to thank: Dr Benillouche, Dr Mourey and Dr Dechoux for their involvement

Conflict of interest:

The authors declared they have no conflict of interests.

Funding:

The study did not receive any funding.

13
References :

[1] Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, et al. Cancer

incidence and mortality worldwide: Sources, methods and major patterns in
GLOBOCAN 2012. Int J Cancer n.d.;136:E359–86. doi:10.1002/ijc.29210.
[2] Colditz GA, Bohlke K. Priorities for the primary prevention of breast cancer. CA Cancer
J Clin n.d.;64:186–94. doi:10.3322/caac.21225.
[3] Coleman C. Early Detection and Screening for Breast Cancer. Semin Oncol Nurs
2017;33:141–55. doi:10.1016/j.soncn.2017.02.009.
[4] Institut National Du Cancer. Résultats du programme de dépistage organisé - Dépistage
du cancer du sein 2017.
[5] HAS. Dépistage du cancer du sein en médecine générale 2013. http://www.has-
sante.fr/portail/jcms/c_272382/fr/depistage-du-cancer-du-sein-en-medecine-
generale?xtmc=&xtcr=112 (accessed November 19, 2013).
[6] Oeffinger KC, Fontham ETH, Etzioni R, Herzig A, Michaelson JS, Shih Y-CT, et al.
Breast Cancer Screening for Women at Average Risk: 2015 Guideline Update From the
American Cancer Society. JAMA 2015;314:1599–614. doi:10.1001/jama.2015.12783.
[7] Canadian Task Force on Preventive Health Care. Breast Cancer—Clinician CBE/BSE
Recommendation n.d. https://canadiantaskforce.ca/breast-cancer-clinician-cbebse-
recommendation/ (accessed July 19, 2018).
[8] Nelson HD, Fu R, Cantor A, Pappas M, Daeges M, Humphrey L. Effectiveness of Breast
Cancer Screening: Systematic Review and Meta-analysis to Update the 2009 U.S.
Preventive Services Task Force Recommendation. Ann Intern Med 2016;164:244–55.
doi:10.7326/M15-0969.
[9] Kösters JP, Gøtzsche PC. Regular self‐examination or clinical examination for early
detection of breast cancer. Cochrane Libr., John Wiley & Sons, Ltd; 2003.
doi:10.1002/14651858.CD003373.
[10] American College of Radiology. Breast imaging reporting and data system—
mammography. Am. Coll. Radiol. Ed Breast Imaging Report. Data Syst. 4th edition
Reston, 2003.
[11] Dilhuydy MH. Breast imaging reporting and data system (BI-RADS) or French
“classification ACR” What tool for what use? A point of view. Eur J Radiol
2007;61:187–91. doi:10.1016/j.ejrad.2006.08.032.
[12] Paty A-C, Ancelle-Park R, Bloch J. Dépistage du cancer du sein Rapport d’évaluation du
suivi épidémiologique 2004.
[13] Chamberlain J, Coleman D, Ellman R. Sensitivity and specificity of screening in the UK
trial of early detection of breast cancer. Cambridge University Press 1992.
[14] Miller AB, Baines CJ, To T, Wall C. Canadian National Breast Screening Study: 2.
Breast cancer detection and death rates among women aged 50 to 59 years. CMAJ Can
Med Assoc J J Assoc Medicale Can 1992;147:1477–88.
[15] Buderer NMF. Statistical Methodology: I. Incorporating the Prevalence of Disease into
the Sample Size Calculation for Sensitivity and Specificity. Acad Emerg Med
1996;3:895–900. doi:10.1111/j.1553-2712.1996.tb03538.x.
[16] Lehman CD, Arao RF, Sprague BL, Lee JM, Buist DSM, Kerlikowske K, et al. National
Performance Benchmarks for Modern Screening Digital Mammography: Update from
the Breast Cancer Surveillance Consortium. Radiology 2016;283:49–58.
doi:10.1148/radiol.2016161174.
[17] Rosenberg RD, Yankaskas BC, Abraham LA, Sickles EA, Lehman CD, Geller BM, et
al. Performance Benchmarks for Screening Mammography. Radiology 2006;241:55–66.
doi:10.1148/radiol.2411051504.

14
[18] Oestreicher N, White E, Lehman CD, Mandelson MT, Porter PL, Taplin SH. Predictors
of sensitivity of clinical breast examination (CBE). Breast Cancer Res Treat
2002;76:73–81.
[19] Walker S, Hyde C, Hamilton W. Risk of breast cancer in symptomatic women in
primary care: a case–control study using electronic records. Br J Gen Pr 2014;64:e788–
93. doi:10.3399/bjgp14X682873.
[20] Paterok EM, Rosenthal H, Säbel M. Nipple discharge and abnormal galactogram.
Results of a long-term study (1964–1990). Eur J Obstet Gynecol 1993;50:227–34.
doi:10.1016/0028-2243(93)90205-Q.
[21] Provencher L, Hogue JC, Desbiens C, Poirier B, Poirier E, Boudreau D, et al. Is clinical
breast examination important for breast cancer detection? Curr Oncol 2016;23:e332–9.
doi:10.3747/co.23.2881.
[22] Lastier D. Programme de dépistage du cancer du sein en France : résultats 2010,
évolutions depuis 2006 2013:26 p.
[23] Romanoff A, Constant TH, Johnson KM, Guadiamos MC, Vega AMB, Zunt J, et al.
Association of Previous Clinical Breast Examination With Reduced Delays and Earlier-
Stage Breast Cancer Diagnosis Among Women in Peru. JAMA Oncol 2017;3:1563–7.
doi:10.1001/jamaoncol.2017.1023.
[24] Dimassi K, Gharsa A, Chanoufi MB, Sfar E, Chelli D. Le traitement conservateur du
cancer du sein: expérience d’une équipe tunisienne. Pan Afr Med J 2014;19.
doi:10.11604/pamj.2014.19.148.4195.
[25] Yen AM-F, Tsau H-S, Fann JC-Y, Chen SL-S, Chiu SY-H, Lee Y-C, et al. Population-
Based Breast Cancer Screening With Risk-Based and Universal Mammography
Screening Compared With Clinical Breast Examination: A Propensity Score Analysis of
1 429 890 Taiwanese Women. JAMA Oncol 2016;2:915–21.
doi:10.1001/jamaoncol.2016.0447.
[26] Gao D, Hu Y, Wang W, Chen F, Pan L, Yuan Y, et al. [Evaluation on the effect of
intervention regarding breast self-examination for decreasing breast cancer mortality].
Zhonghua Liu Xing Bing Xue Za Zhi Zhonghua Liuxingbingxue Zazhi 2006;27:985–90.
[27] Hassan LM, Mahmoud N, Miller AB, Iraj H, Mohsen M, Majid J, et al. Evaluation of
effect of self-examination and physical examination on breast cancer. The Breast
2015;24:487–90. doi:10.1016/j.breast.2015.04.011.
[28] Paci E, EUROSCREEN Working Group. Summary of the evidence of breast cancer
service screening outcomes in Europe and first estimate of the benefit and harm balance
sheet. J Med Screen 2012;19 Suppl 1:5–13. doi:10.1258/jms.2012.012077.
[29] National Health Services. Be breast aware 2006.
[30] Institut national du cancer. Plan d’action pour la rénovation du DO en France 2017.
[31] World Cancer Reasearch Fund, American Institute for Cancer Research. Food, Nutrition,
Physical Activity, and the Prevention of Cancer, a Global Perspective 2007.
[32] Duclos M. Effet de l’exercice sur le cancer. Medscape 2012.
http://francais.medscape.com/viewarticle/3372703 (accessed March 14, 2018).
[33] Seven M, Akyüz A, Robertson LB. Interventional Education Methods for Increasing
Women’s Participation in Breast Cancer Screening Program. J Cancer Educ Off J Am
Assoc Cancer Educ 2015;30:244–52. doi:10.1007/s13187-014-0709-8.
[34] O’Mahony M, Comber H, Fitzgerald T, Corrigan MA, Fitzgerald E, Grunfeld EA, et al.
Interventions for raising breast cancer awareness in women. Cochrane Database Syst
Rev 2017. doi:10.1002/14651858.CD011396.pub2.
[35] Weigert J, Cavanaugh N, Ju T. Evaluating Mammographer Acceptance of MammoRisk
Software. Radiol Technol 2018;89:344–50.

15
[36] Chowdhury S, Dent T, Pashayan N, Hall A, Lyratzopoulos G, Hallowell N, et al.
Incorporating genomics into breast and prostate cancer screening: assessing the
implications. Genet Med 2013;15:423–32. doi:10.1038/gim.2012.167.

16
 Figure 1. Flow chart describing the patients and their clinical abnormalities

3557 patients screened

339 patients with personal
history of breast cancer

3218 included patients

2505 patients with « normal » 713 patients with « abnormal »

clinical breast examination clinical breast examination*
(77.8%) (22.2%)

Bi-RADS 4/5:
85 (3.4%)

Breast mass: Breast pain: Nipple Axillary Retraction: Other:

469 (14.6%) 126 (3.9%) discharge: nodes: 9 (0.3%) 169 (23.7%)
16 (0.5%) 17 (0.5%)

Bi-RADS 4/5: Bi-RADS 4/5: Bi-RADS 4/5: Bi-RADS 4/5: Bi-RADS 4/5: Bi-RADS 4/5:
34 (7.2%) 6 (4.8%) 5 (31.2%) 3 (17.6%) 2 (22.2%) 5 (3.0%)

*Patients could have more than one abnormality

1
 Table 1. Characteristics of patients depending on the mammography result

Overall population High risk Lower risk p

N = 3218 N = 133 N = 3085
n(%) n(%) n(%)
Patients characteristics
Age mean
55.1 (10.4) 53.8(12.4) 55.2 (10.3) 0.15
(standard deviation)
Organised screening 512 (15.9%) 18 (13.5%) 494(16.0%) 0.52
Familial history of breast
1285 (39.9%) 48 (36.1%) 1237 (40.1%) 0.40
cancer
History of benign breast
420 (13.1%) 26 (19.5%) 394 (12.8%) 0.03
neoplasm
Nulliparous 885 (27.5%) 46 (34.8%) 839 (27.3%) 0.07
Clinical breast examination

Any clinical abnormalitya 713 (22.2%) 48 (36.1%) 665 (21.6%) <0.01

≥ 2 clinical abnormalitiesa,b 88 (2.7%) 8 (6.0%) 78 (2.5%) 0.02
Breast mass 469 (14.6%) 34(25.6%) 435 (14.1%) <0.01
Pain 126 (3.9%) 6 (4.5%) 102 (3.9%) 0.64
Nipple discharge 16 (0.5%) 5 (3.8%) 11 (0.4%) <0.01
Axillary nodes 17 (0.5%) 3 (2.3%) 14 (0.5%) 0.03
Retraction 9 (0.3%) 2 (1.5%) 7 (0.2%) 0.05
Erythema 5 (0.2%) 0 5 (0.2%) 1
Œdema 2 (0.1%) 2 (1.5%) 0 <0.01
Other 116 (3.6%) 3 (2.3%) 113 (3.7%) 0.63
a
Clinical abnormalities could be: mass, retraction, oedema, erythema, orange skin, ulceration, pain,
nipple discharge, axillary node, other (including breast erythema, breast oedema, hematoma, varicose
veins)
b
Among these 88 women, 2 patients had 3 abnormalities and had a high risk mammography

1
Table 2. Accuracy of clinical breast examination for the overall population

N=3218 Se Sp PPV NPV

n (%) [95%CI] [95%CI] [95%CI] [95%CI]
Any clinical 36% 78% 7% 97%
713 (22%)
abnormality* [28%; 45%] [77%; 80%] [5%; 9%] [96%; 97%]
≥ 2 clinical 6% 97% 9% 96%
86 (3%)
abnormalities* [3%; 12%] [97%; 98%] [4%; 18%] [95%; 97%]
26% 86% 7% 96%
Breast mass 469 (15%)
[18%; 34%] [85%; 87%] [5%; 10%] [96%; 97%]
2% 100% 22% 96%
Retraction 9 (0.2%)
[0%; 5%] [100%; 100%] [3%; 60%] [95%; 97%]
2% 100% 100% 96%
Oedema 2 (0.1%)
[0%; 5%] [100%; 100%] [9%; 100%] [95%; 97%]
0% 100% 0% 96%
Erythema 5 (0.1%)
[0%; 4%] [100%; 100%] [0%; 64%] [95%; 97%]
5% 96% 5% 96%
Pain 126 (4%)
[2%; 10%] [95%; 97%] [2%; 10%] [95%; 97%]
Nipple 4% 100% 31% 96%
16 (0.5%)
discharge [1%; 9%] [99%; 100%] [11%; 59%] [95%; 97%]
2% 100% 18% 96%
Axillary node 17 (0.5%)
[0%; 6%] [99%; 100%] [4%; 43%] [95%; 97%]
2% 96% 3% 96%
Other 116 (4%)
[0%; 6%] [96%; 97%] [1%; 7%] [96%; 96%]
* Clinical abnormalities could be: mass, retraction, oedema, erythema, orange skin, ulceration, pain, nipple
discharge, axillary node, other
(Se: sensitivity, Sp: specificity, PPV: positive predictive value, NPV: negative predictive value, CI: confidence
interval)

2
Table 3. Accuracy of clinical breast examination for patients under 50 years

N=957 Se Sp PPV NPV

n(%) [95%CI] [95%CI] [95%CI] [95%CI]
Any clinical 38% 73% 6% 96%
262 (27%)
abnormality* [24%; 53%] [70%; 76%] [4%; 10%] [94%; 97%]
≥ 2 clinical 4% 96% 5% 95%
41 (4%)
abnormalities* [1%; 15%] [94%; 97%] [1%; 17%] [94%; 97%]
27% 83% 7% 96%
Breast mass 170 (18%)
[15%; 42%] [80%; 85%] [4%; 12%] [94%; 97%]
2% 94% 2% 95%
Pain 53 (6%)
[0%; 12%] [93%; 96%] [0%; 10%] [94%; 96%]
Nipple 7% 99% 27% 96%
11(1%)
discharge [1%; 18%] [98%; 100%] [6%; 61%] [94%; 97%]
4% 99% 29% 95%
Axillary node 7 (1%)
[1%; 15%] [99%; 100%] [4%; 71%] [94%; 97%]
* Clinical abnormalities could be: mass, retraction, oedema, erythema, orange skin, ulceration, pain, nipple
discharge, axillary node, other
(Se: sensitivity, Sp: specificity, PPV: positive predictive value, NPV: negative predictive value, CI: confidence
interval)

3
Table 4. Accuracy of clinical breast examination for patients between 50 and 74 years.

N=2190 Se Sp PPV NPV

* n(%) [95%CI] [95%CI] [95%CI] [95%CI]

Any clinical 36% 80% 7% 97%

441 (20%)
abnormality* [25%; 47%] [79%; 82%] [4%; 9%] [96%; 98%]

≥ 2 clinical 7% 98% 14% 97%

44 (2%)
abnormalities* [3%; 15%] [98%; 99%] [5%; 27%] [96%; 97%]

25% 87%
7% 97%
Breast mass 290 (13 %) [16%; 36%] [86%; 89%]
[4%; 10%] [96%; 98%]

2% 100% 40% 96%

Retraction 5(0.2%)
[0%; 9%] [100%; 100%] [5%; 85%] [96%; 97%]

6% 97% 7% 96%
Pain 73 (3%)
[2%; 14%] [96%; 97%] [2%; 15%] [96%; 97%]

Nipple 2% 100% 40% 96%

5(0.2%)
discharge [0%; 9%] [100%; 100%] [5%; 85%] [96%; 97%]

1% 100% 10% 96%

Axillary node 10 (0.5%)
[0%; 7%] [99%; 100%] [0%; 45%] [95%; 97%]

Clinical abnormalities could be: mass, retraction, oedema, erythema, orange skin, ulceration, pain, nipple
discharge, axillary node, other

(Se: sensitivity, Sp: specificity, PPV: positive predictive value, NPV: negative predictive value, CI: confidence
interval)