With the wide implementation of breast cancer imaging screening, an increasing number of small-sized breast cancers are being detected (1–3). Women with small breast cancer generally have favorable long-term outcomes (3–5). Nevertheless, breast cancer is a heterogeneous disease with complicated pathology and biological behavior, and some women eventually die of breast cancer or have early recurrence despite having a small tumor (6, 7). Indeed, tumor size by itself does not tell much about the biological behavior because a specific tumor size can result from an indolent disease that has been growing for some time or from a very aggressive disease that has been developing for a few weeks (5, 6, 8–10). In addition, even though tumor size is a well-recognized prognostic factor for invasive breast cancer, large ductal carcinoma in situ (DCIS) carries the risk of harboring microinvasive foci that can lead to poor outcomes (11).

Despite their value in predicting the long-term outcomes, traditional prognostic factors (such as pathological grade and lymph node status) and the more powerful molecular and genetic examinations might not be sufficient for the prognosis of small breast cancer (12–16). Some women with small breast cancer still die from breast cancer (6, 7). Clinicopathological factors like pathological tumor grade and lymph node status are commonly used for breast cancer prognosis (17, 18) but are not as effective in small breast cancers as in larger ones (12–16). Thus, other reliable prognostic factors to identify the more aggressive subset among small breast cancers as early as possible are needed to ensure appropriate management.

Breast ultrasound features may reflect the underlying pathological manifestations and biological behavior of a breast tumor, which would be reliable imaging indicators for breast cancer outcomes (15, 19). Still, most studies focused on assessing the associations between ultrasound features and breast cancer risk categories like molecular type and histological grade (16–18). In contrast, studies examining the direct associations between ultrasound features and breast cancer mortality are insufficient, especially for small breast cancers <20 mm.

Therefore, this study aimed to explore whether ultrasound features could identify small breast cancers with poor outcomes.

This study aimed to explore whether ultrasound features could identify small breast cancers with poor outcomes. The results suggested that spiculated morphology and anti-parallel orientation at ultrasound were associated with poor BCSS and DFS in patients with breast cancer <20 mm.

The BCSS in the present study was 96.2%, which is consistent with previous studies in similar study populations (3–5). In the present study, lymph node metastasis was a statistically significant predictor for small cancer. While no significant differences were observed regarding tumor grade. Among the 30 breast cancer deaths, only 10 patients were diagnosed as grade 3 tumors. Grade 3 tumors only accounted for 24.8% (196/790) among the small breast cancers including in this study, which is consistent with the theory that pathological grade tend to progress towards a higher grade during breast cancer growth (22). This biological characteristic might explain the different value of tumor grade in predicting outcome among breast cancers with different sizes. It was reported that the involvement of lymph nodes is strongly related to the metastatic potential of breast cancer (22), and lymph node involvement in early stage, small breast cancers might reflect a more aggressive biological behavior (5, 23). In this study, positive lymph node metastases were associated with poor BCSS in tumors of 10-20 mm, but not in tumors of 1-10 mm, as few positive lymph nodes were detected among tumors of 1-10 mm (n=20), which is supported by Tabar et al. (13). Hence, the traditional prognostic factors might not be reliable in breast cancers <20 mm.

Ultrasound imaging, which reflects the underlying pathological and biological characteristics of the breast tumors, might be a complementary and reliable way to characterize small breast cancers. Previous studies (13, 14) identified the role of mammography features in identifying high-risk small breast cancers. The association between ultrasound features and triple-negative breast cancer death was examined, but only 50.6% were tumor <20 mm (24). Chung et al. (25) found no association between ultrasound features and outcome by dividing tumors into mass or non-mass without detailed sonography features. Still, research concerning the predictive value of sonographic findings for long-term outcomes for small tumors is insufficient. In this study, spiculated morphologyand anti-parallel orientation, alone or in combination, were related to cancer death and recurrence, and the differences remained significant after adjusting for age, tumor grade, lymph node status, adjuvant therapy, and molecular biomarker. Survival to breast cancer <20 mm is favorable except for tumors with spiculated morphologycombined with a anti-parallel orientation. It is proposed that a non-proportional growth is a suspicious indicator for malignancy and poor prognosis (19, 24, 26), which supports the results of the present study. Compared with ductal carcinoma in situ, invasive ductal carcinoma is more likely to have a non-proportional growth orientation (27). Guo et al. (19) assessed the value of ultrasound features in 336 breast cancers and supposed that vertical growth tends to correlate with high-risk breast cancer. At the pathological examination, indolent cancers generally show a homogeneous expansive growth, while a non-proportional growth at ultrasound reflects a heterogeneous infiltrating growth of malignant cells through lobule of mammary gland boundaries (24, 28, 29) and indicates a more aggressive biological behavior, particularly when it occurs early in breast cancer development.

In this study, the presence of spiculated morphology was another factor associated with poor prognosis. Results about the correlation between spiculated morphology and breast cancer prognosis remain conflicting. Wu et al. (30) proposed that spiculated margins are significantly related to positive HER2, a factor for high tumor grade and poor prognosis (31). While some studies consider spiculated morphology are more frequently associated with the low-grade tumor and positive ER and PR (21, 32). The conflicting results might be contributed to the analysis of breast cancers of different sizes. Besides, the long-term outcomes of small breast cancer were observed, and the association of spiculated morphology and cancer-related mortality was identified in this study, while most previous studies compared the imaging findings between different tumor attributes, including pathology characteristics and molecular biomarkers, which might explain the inconsistence. In terms of pathological manifestations, a spiculated margin is a reflection of the invasive growth of malignant cells into the normal surrounding tissues, indicating a high invasive capability (30, 33, 34). In addition, various growth factors (including the vascular endothelial growth factor and transforming growth factor) can be secreted by malignant cells to stimulate angiogenesis, infiltration, and metastasis (34, 35). An infiltrating growth early in tumor development might represent a more aggressive biological behavior and a graver prognosis, which was consistent to our results.

Casting-type calcification in mammography is a reliable predictor for a worse prognosis for small breast cancer (14), but no significant associations with BCSS and DFS were observed in the present study. It might be because of the sensitivity of ultrasound for calcifications. Molecular biomarkers are widely used in the clinical decision-making of breast cancer management, but no significant associations were observed between the biomarkers (ER, PR, HER2, and Ki67) and breast cancer outcomes in this study. Tryfonidis et al. (36) reported that small triple-negative and HER2-positive breast cancers tend to have a worse prognosis, while Galimberti et al. (37) and Bao et al. (7) proposed that there are no correlations between the status of ER, PR, HER2, and the outcome events. We are cautious about the results of the role of the molecular biomarker in breast cancer prognosis because of different cut-off values and a median follow-up time of only 3.5 years in this study, which might not be enough to observe the effects of molecular biomarkers on the outcomes of small breast cancer.

The treatment of each breast cancer is based on the careful consideration of recognized prognostic factors like tumor size, tumor grade, lymph node status, and molecular subtype. However, this decision-making assessment for small breast cancers might not be the same as for larger ones. Among the 30 breast cancer deaths in this study, 13 had negative lymph nodes, and 19 had grade 1-2 tumors, and one patient was even diagnosed with ductal carcinoma in situ at the final pathological examination. The combination of spiculated morphology and anti-parallel orientation was an independent predictor for breast cancer death in this study. Combining the imaging findings with the existing prognosis factors could contribute to the clinical decision-making by identifying the small breast cancers with a poor prognosis, for which more aggressive treatments might be warranted. Ultrasound is a readily available and inexpensive imaging modality that can be used alone or in combination with mammography. In a developing country where healthcare resources can be scarce in specific regions, standalone ultrasound can be used for breast cancer screening since some ultrasound systems are portable and mammography and magnetic resonance imaging might be unavailable.

There are some limitations to this study. First, it was a retrospective and single-center study, and bias might exist. Second, the classification of different ultrasound features is subjective and operator-dependent. Experienced radiologists are required to obtain reliable results. A high image resolution is required to observe the spiculated in small breast cancer. Third, the follow-up time was not long enough, and the number of breast cancer-specific death was small. Finally, as the number of deaths from cancer in the series is very low, there are very few findings that could be significant. Further studies are required.

The combined presence of spiculated morphology and anti-parallel orientation in breast ultrasound might be associated with poor BCSS and DFS in patients with primary breast cancers <20 mm. Combining these imaging findings as complementary risk indicators with existing factors might improve the clinical evaluation in small breast cancers.

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding author.

This retrospective study was approved by the institutional ethics committee of Shanghai General Hospital. Written informed consent was waived by the institutional review board.

Conceptualization: RW; Data curation: SHS; Formal analysis: SHS; Funding acquisition: RW; Investigation: CXL; Methodology: YZ; Project administration: MHY; Resources: XL; Software: JC; Supervision: RW; Validation: JFW; Roles/writing - original draft: SHS; Writing - review and editing: MHY.