Introduction

With improvements in breast imaging, the role of clinical breast exam (CBE) in screening average risk women is increasingly being questioned; with a recent report suggesting a very low yield in women undergoing regular mammography screening [1]. The American Cancer Society (ACS) does not recommend CBE for BC screening in average risk women at any age [2]. This recommendation is based on lack of level 1 evidence of any proven benefit for CBE either as a stand-alone tool or in conjunction with screening mammography [3]. The National Cancer Comprehensive Network [4] recommends a clinical encounter starting at age 25, which includes a CBE in asymptomatic individuals. Despite lack of proof of efficacy of CBE in average risk women, it may play a role in high-risk women, specifically BRCA mutation carriers. Inclusion of CBE in high-risk surveillance protocols varies worldwide, with some societies recommending annual or bi-annual CBE, while others do not [5, 6]. The ESMO recent recommendations [7] for surveillance of BRCA mutation carriers clearly state that “Clinical breast examination is of no value as a screening tool”, referencing a report by Hettipathirana et al [1.

Fig. 1
figure 1

1 Including breast cancers that were diagnosed prior to the study and therefore excluded. 2 Incidental- the clinical abnormality was not at the location of the cancer

Flow chart of study cohorts.

Full size image

Six (4.5%) women presented with interval cancers, median age was 36 years (range 28–67). Median time from previous imaging was 3 months (range 0.11–0.66). Time from previous MRI ranged between 7.5 months and 3 years; one woman had never had a previous MRI. Average tumor size was 18 mm (range 10–46mm). Four were node negative, one was node positive, and for one this data was missing. Two were triple negative, two were luminal, one was HER2neu positive and data were missing for the last.

Three cancers (2%; 95% CI 0.7–6) were diagnosed secondary to CBE findings, none of these women had an MRI during this round (Table 2). In the screening rounds not including MRI three of the BCs were diagnosed by CBE (7.7%; 95% CI 2.6–20.3).

In the Health screening center 7,949 average to intermediate risk women had 15,518 CBEs done during the study period. The final cohort included 87 women diagnosed with BC after a visit to the clinic with a documented CBE within 6 months of the diagnosis (Fig. 1).

Most of these women (54; 62%) had no known family history of BC (Table 1). Four women with no known family history were subsequently identified as BRCA mutation carriers.

Table 1 Comparison of average to intermediate risk group with BRCA mutation carrier group
Full size table

The imaging, pathology and treatment characteristics are summarized in Table 1.

The median age of women presenting with an interval cancer was 64.3 years (range 41–81). Median time from previous imaging was 0.8 years (range 0.73–10).

There were 3 (4%; 95% CI 1–9) cancers that were diagnosed secondary to an abnormal CBE (Table 2). One was found on MRI; the other 2 on US. In two of these cases, the palpable finding was in a different location from the cancer diagnosed. Nonetheless, the CBE initiated the work-up.

Table 2 Cancers diagnosed secondary to abnormal CBE
Full size table

As expected, when compared to the average to intermediate risk group, BRCA mutation carriers were diagnosed with BC at an earlier age; mainly by MRI; and more often with invasive, high grade triple negative cancers. They underwent more often mastectomies and received systemic chemotherapy (Table 1).

In both groups, diagnosis of cancer by CBE only was a rare event; over 4,000 exams were needed in order to diagnose one BC.

Discussion

In order to assess the yield of CBE in high-risk women we compared the additional cancer yield of CBE in BRCA carriers to average to intermediate risk women undergoing regular screening mammography. In the current study, the additional yield of CBE to BC diagnosis in both BRCA mutation carriers and in average to intermediate risk population was marginal at best. Notably, the number of CBE needed in order to find one breast cancer was over 4,000 in BRCA mutation carriers. These results are in line with previous studies in which the additional yield of CBE to BC detection ranged from 0 to 6% (Table 3). The incorporation of breast MRI in the surveillance protocol of BRCA mutation carriers resulted in a decrease in the proportion of women presenting with interval cancers from 35–50% [9] to 0–19% [10,11,12,13, 15, 16] (Table 3).

Table 3 Studies reporting cancers detected in BRCA mutation carriers by CBE
Full size table

In 2014, Roeke et al. [17] systematically reviewed the additional cancer yield of CBE in women at increased risk of BC, and reported that it ranged between 0 and 4% in 7 prospective studies. The recent ESMO guidelines removed the previous recommendation for CBE in surveillance of BRCA mutation carriers, while recommending in BRCA1 mutation carriers imaging every 6 months preferably by MRI [7]. The National Institute of Health and Care Excellence (NICE) continues to recommend breast awareness and annual MRI [18], while NCCN continues to recommend CBE from age 25 and annual MRI [19].

Based on the results of the current study, and given the well- established superiority of MRI over other breast imaging modalities [20], combined with the fact that all 3 BC cases in BRCA mutation carriers identified by CBE were in women not undergoing MRI at the same screening round, it appears safe to forgo the CBE at the time surveillance MRI is performed.

This study has several limitations. Data on visits to the high-risk clinic and the health screening center were based on computer queries and therefore errors in coding may have resulted in inaccurate estimation of the total number of women and of the total number of visits to these clinics. Although charts of women subsequently diagnosed with breast cancer were manually reviewed, the total number of women and of visits was based on retrieval of visit codes. This may impact the accuracy of the estimation of the number needed to screen in order to detect one BC. As the estimates of the additional cancer yield of CBE are based on women diagnosed with breast cancer, these estimates are minimally affected by coding errors. The BRCA cohort included women after a previous diagnosis of BC (as these women are recommended to continue the same surveillance scheme), whereas we excluded women with a history of BC in the average to intermediate risk group. Women that underwent bilateral mastectomy were included in the BRCA cohort despite controversy regarding their continued need for increased surveillance. With the increasing prevalence of risk-reducing mastectomy, CBE may play a bigger role in the surveillance of these women, however our numbers are too small to reach meaningful conclusions. The surgeons performing the CBEs were a heterogenous group, with varying experience and abilities. CBE is a skill which is almost impossible to quantify and analyze objectively. The population analyzed herein is from a single medical center in Israel and may not reflect the reality in other medical centers in the country. The compliance of average risk women to BC screening is below 80% in Israel and this may have impacted the results as well.

In summary, based on a large population of BRCA mutation carriers, it appears that CBE has a marginal contribution to the diagnosis of BC, specifically during the screening round that does not include MRI. It seems safe not to perform CBE during the screening visit that includes an MRI. In average to intermediate risk women undergoing regular BC screening the yield of CBE is very low, and may not be justified at all.