Small BC not only has relatively good prognosis, but also increases the possibility of breast conservation. According to TNM classification of the AJCC, small tumors are all considered as T1mi, a, b, or c tumors[9]. T1mi is a tumor that is 1mm or smaller. One study defined small tumors as tumors with tumor size ≤ 10mm (T1a, T1b)[16] and another study defined a group of small tumors by combining small diameter (≤ 10 mm, T1ab tumors) with low tumor cell proliferation (≤ 10% Ki67 expression rate)[17]. From the perspective of imaging examinations, screening BC less than 10mm is still difficult. Presently, enhanced MRI is the most reliable method to find small breast tumors when diagnosing BC[18]. US and mammography are still used to screen BC because of the long duration and high costs of MRI[19]. Although US has advantages in detecting BC nodules, previous research on small nodules with less than or equal to 10mm on US and clear BC pathology results are still insufficient[20]. In view of this, we chose to evaluate the significance of US and mammography screening using nodules with less than or equal to 10mm on US.
Our results indicate that the size of BC diagnosed by US may not be completely consistent with pathological results. The most accurate size of tumor staging should be the pathological size. Imaging examination results can only be extremely close but cannot replace pathological examination results. Many studies that evaluate the accuracy of tumor size using mammography, US, and MRI by comparing them with pathological size find that these imaging methods can either under- or over-estimate tumor size[21, 22]. Of the 491 BC patients with tumor size less than or equal to 10mm, 401 (81.7%) patients’ imaging results were consistent with pathological sizes (less than or equal to 10mm). For invasive cancer, US size results show little difference from the pathological size results; for in situ cancer that cannot be shown as a nodule or is diffusely displayed pathologically, US size values tend to be larger than pathological values. This result is consistent with the results reported in a previous literature[23], although the size range of BC selected in said previous literature is larger (4.5-38.1mm) and not limited to 10mm or less[23]. Other previous studies have found that imaging significantly underestimates the size of DCIS[24]. Even mammography is more accurate than US in measuring the maximum cancer diameter in DCIS of BC[25]. This may be why the 56 (62.2%) out of the 90 patients in our study who had inconsistent US and pathological results were diagnosed with CIS.
Mammography and US should complement one another[26]. In this study, we found that mammography had advantages in the diagnosis of CIS based on the higher sensitivity and specificity of mammography in displaying fine calcification than US. According to previous literature reports, the morphology of calcification may indicate the possibility of cancer tissue [27]. BI-RADS classification is carried out by radiologists based on these characteristics. In the diagnosis of CIS, mammography has irreplaceable diagnostic value based on fine calcification. Mammography does have its own limitations. Previous studies have reported that Mammography has low sensitivity to small invasive breast cancer less than 15mm[28]. This result is consistent with our research findings. There is a fair number of false-negative mammograms[29]. Bae, M S et al. found that 81% of BC detected at screening US cannot be seen at all in mammography[30]. A study on Chinese women found that US is more sensitive and accurate than mammography[31]. Therefore, it is worth noting that US should not be replaced by mammography.
US has more advantages in diagnosing nodular invasive cancer[32], even when the tumor size is very small. In our study, a higher proportion of invasive cancer appears on US with unclear margins, irregular shapes, more color flow; in particular, the higher proportion of color flow aligns with a previous study that found correlation between vascularization and breast cancer clarification[33]. It is consistent with a previous report that US can categorize BC into US-BI RADS categories according to sonogram characteristics such as margins, shapes and color flow.
Nodular invasive cancer such as breast neuroendocrine carcinoma and mucinous carcinoma displays certain special characteristics on US - breast neuroendocrine carcinoma shows more irregular nodular changes without calcification and mucinous carcinoma shows more color flow. These characteristics gives US an advantage in their detection. In our study, cases of both types of carcinomas had 100% diagnostic rates (see Table 2), which may be evidence that the signature characteristics of these carcinomas make US an especially accurate method. However, it is worth noting that our study may not have enough cases to accurately represent diagnosis for these two types of carcinomas. US display of DCIS is not as good as that of invasive cancer due to the non-mass appearance of DCIS[34]. This also allows mammography, which is sensitive to fine calcification in CIS, to supplement the shortcomings of US.
The detection of small high-grade invasive cancers is vital to reducing BC mortality. The mortality rate of BC has declined significantly in the past years. On the one hand, it is thanks to the progress of treatment level, yet on the other hand, it is thanks to early screening. According to 4 simulation models in a previous study, compared with interventions in 1975, BC screening and treatment in 2019 were associated with a 58% reduction in US BC mortality[35].
Previous literature has reported that multiple clinicopathological factors are associated with recurrence and metastasis of BC. For invasive breast cancer, factors associated with recurrence were age younger than 40 years, the triple-negative subtype, and BI-RADS category 4A lesions[36]. Locoregional recurrence occurred with different patterns according to BC subtypes, with younger patients having greater differences in patterns among subtypes than older patients[37]. Tumors with the same anatomic stage were assigned different prognostic stages depending on the histologic grade, hormone receptor status, HER2 status, and multigene panels[38]. In our study, women aged 40 years or younger or with pathological results showing invasive carcinoma had higher hazard rates for recurrence and metastasis than those older than 40 years or with CIS and/or microinvasion. This result is partially consistent with previous research findings above.
There were several limitations to this study. Firstly, since this is a retrospective analysis of large-scale breast screenings, it must involve more than one breast US doctor. Although US-RADS provides a standard for analyzing US images and the participating breast US doctors are all senior and experienced US doctors, individual differences in analysis are still inevitable. Secondly, although there are relatively standardized treatment methods in this research process, there may still be individual differences in treatment during the study period that may affect prognostic factors. These factors may differ from the factors chosen in other reports. Moreover, this study did not include sufficient cases with long-term (> 120month years) follow-up. It is also worth noting that, in addition to traditional US, multimodal imaging such as contrast-enhanced US, elasticity, and even ABVS can provide more information to accurately diagnose BC. These additional imaging methods may also help prognosis and could be reflected in future research.