Pathological and radiological assessment of benign breast lesions with BIRADS IVc/V subtypes. should we repeat the biopsy?

Abstract

Background

Timely diagnosis is a crucial factor in decreasing the death rate of patients with breast cancer. BI-RADS categories IVc and V indicate a strong suspicion of cancer. The categorisation of each group is determined by the characteristics of the lesion. Certain benign breast lesions might have radiological features indicative of malignancy; thus, biopsy is mandatory. This study aimed to identify the histopathological diagnosis of benign breast masses classified into BIRADS IVc and V subgroups, investigate the radiological characteristics of these masses, and identify ultrasound features that could lead to false positive results (benign lesions that mimic malignancy on imaging).

Methods

This was a retrospective cross-sectional study at a single facility. Breast lesions reported as BIRADS IVc and V that underwent needle core/stereotactic vacuum-assisted biopsy were reviewed. Patients with benign pathologic diagnoses were analysed, delineating pathological diagnoses. Radiological descriptors were compared to those of a matched control of 50 malignant cases with BIRADS IVc.

Results

A total of 828 breast lesions classified as BIRADS IVc or V were detected during the period spanning from 2015 to 2022. Forty-four lesions (44/828, 5.3%) were benign at initial biopsy, while 784 lesions (784/828, 94.7%) were malignant. After histopathological testing and repeat biopsy, 26/828 (3.14%) patients had discordant benign diagnosis. Half of the repeated biopsies (10/20, 50%) showed malignant pathology. Compared to that in the control group, the presence of an oval shape of the mass was significantly more common in patients with benign pathology (p = 0.035). Conversely, the presence of posterior shadowing was significantly less common (p = 0.050) in benign lesions. No significant differences were observed for the other radiological characteristics. The most common histopathological diagnosis was fibrocystic change.

Conclusion

This study highlights key findings regarding the sonographic imaging descriptors and histopathological diagnoses of benign breast lesions categorised as BIRADS IVc/V. The study recommends a correlation between clinical and radiological findings and encourages multidisciplinary decision-making among radiologists, pathologists, and clinicians to determine if a repeat biopsy is warranted. There is a need for continuous research to improve the diagnosis and treatment of breast lesions and reduce false-positive rates by incorporating other methodologies such as sonoelastography and incorporating deep learning and artificial intelligence in the decision-making to eliminate unnecessary procedures.

Background

According to research, breast cancer is the predominant form of cancer among women worldwide [1]. The World Health Organisation (WHO) has emphasised that timely diagnosis is a crucial factor in decreasing the death rate of patients with breast cancer [2]. Ultrasound-guided breast biopsy is a frequently conducted procedure used to diagnose breast lesions. The Breast Imaging Reporting and Data System (BI-RADS) is utilised to categorise breast abnormalities depending on their likelihood of being cancerous, determined by distinct characteristics observed on mammography, ultrasonography, and magnetic resonance imaging of the breast. Category IVc indicates a strong suspicion of cancer and carries a chance of malignancy ranging from above 50% to less than 95% [3]. The BIRADS IVc categorises lesions that have an intermediate level of risk for malignancy but do not exhibit conventional characteristics. The primary radiological characteristic BIRADS IVc consists of solid, irregular masses with poorly defined boundaries or a recent accumulation of microcalcifications exhibiting pleomorphic and delicate calcifications. A BIRADS V indicates a strong possibility of cancer, with a probability of malignancy exceeding 95% [4, 5]. The categorisation of each group is determined by the characteristics of the lesion, such as breast composition/density, masses, calcifications, asymmetries, related abnormalities, and location of the lesion.

While these categories strongly indicate the presence of cancer, a few of them are determined to be noncancerous. A study revealed that the majority of BI-RADS IV lesions exhibited fibrocystic alterations upon histological evaluation, with ductal carcinoma in situ (DCIS) being the second most prevalent finding [6]. The primary objective of the BIRADS system is to establish a uniform method for assessing the likelihood of breast lesions being cancerous. Previous studies conducted by Baker et al. [7] and Berg et al. [8] have highlighted the significant variability in lesion description and therapy.

Research has indicated that the diagnostic accuracy of subcategories IVb and IVc surpasses that of subcategory IVa. The PPVs for subcategories IVa-IVc were considerably influenced by the age of the patient [9]. Employing the subcategories of category IV is a rational approach, irrespective of the proficiency and duration of application by radiologists [9].

This study aimed to identify histopathological diagnoses of patients with breast lesions classified as BIRADS IVc or V, based on biopsy-proven benign diagnosis. It also investigated if repeat biopsy is needed or if the original biopsy results are reliable. The study aimed to determine if certain radiological features are more likely in benign lesions that mimic malignancy, aiding radiologists and clinicians in decision- making. This determination can also reflect the progress of mammographic screening [10].

Rationale of the study:

BIRADS IVc and V classifications are associated with a higher risk of cancer; however, some lesions are found to be benign upon histological testing. This discordance raises important clinical questions: Are the biopsy results conclusive in such cases? Could a repeat biopsy enhance the diagnostic confidence, or does it result in an unnecessary intervention?

Patients and methods

This retrospective, cross-sectional, single-centre study was conducted at a tertiary hospital. The study was approved by the institutional ethics review board, and informed consent was waived since this study was conducted retrospectively, without using any personal patient information. The outcomes did not have any impact on patient treatment. The patient population included all patients with breast lesions reported as BIRAD IVc or V by imaging (breast ultrasound) who underwent ultrasound (US)-guided needle core biopsy or stereotactic vacuum-assisted biopsy and who were reported to have a “nonmalignant” diagnosis after histopathology examination during the period spanning from 2015 to 2022. A matched control group for age and sex is used to assure that these confounding factors do not affect the relationship between the condition and radiological findings. Breast density varies across men and women, as well as across age groups. Additionally, matching can enhance the precision of estimates with smaller sample sizes. More controls were recruited to ensure adequate comparison.

Population

We reviewed 828 patients with BIRAD category IVc or V breast masses who underwent US-guided needle core biopsy or stereotactic vacuum-assisted biopsy. Forty-four patients (44/282, 5.31%) were reported to have a benign diagnosis after histopathological examination between 2015 and 2022. Biopsies were performed by a consultant breast radiologist or a consultant interventional radiologist with 5–10 years of experience. Cases were all females with a mean age of 52.7 years, a maximum age of 91 years, and a minimum age of 26 years. The patients were compared to an age- and sex-matched control group comprising 50 individuals with breast masses classified as BI-RADS IVc on imaging, all of whom were confirmed to have malignancies following ultrasound-guided needle core biopsy conducted during the same period. The control group consisted entirely of females, with a mean age of 52.7 years, a maximum age of 92 years, and a minimum age of 28 years. Radiology reports were retrieved for these patients, and all radiological findings described in the report were collected (laterality, posterior features, echogenicity, margin, shape, and size). In specific cases (when the findings in the report were not adequately described), two radiologists reviewed the images to collect the missing findings, and they were blinded to the final pathology results. The reviewing radiologists included a consultant breast radiologist and a breast imaging fellow. Afterwards, pathology reports were reviewed for final diagnosis and biopsy size.

Data analysis

Chi-square analysis and binary logistic regression analysis using IBM SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA) were performed to determine the associations between radiological and pathological findings and different variables. The statistical significance of each parameter was determined using a 95% confidence interval (CI), and a p value ≤ 0.05 was considered to indicate statistical significance.

Results

A total of 828 breast lesions classified as BIRADS IVc or V were detected in a single medical facility during the period spanning from 2015 to 2022. As summarized in Fig. 1, seven hundred eighty-four cases (94.6%) were initially identified as malignant by the first biopsy, whereas 44 cases (5.31%) were shown to be benign after histological testing. A total of forty patients underwent needle core biopsy guided by ultrasound, whereas four patients underwent stereotactic vacuum-assisted biopsy.

Fig. 1

Study population selection workflow. Hierarchy explaining the study population selection of patients with confirmed benign breast masses with BIRADS IVc and V subtypes

Of the 44 instances of breast lesions, 34/44 (77%) were classified as BIRADS IVc, whereas 10/44 (23%) were classified as BIRADS V.

Thirty-six cases were classified as “mass,” while the remaining 8 cases were classified as “calcifications” (non-mass). Non-mass cases were omitted due to the inapplicability of the radiological parameters studied to calcifications alone.

Upon repeat biopsy of the patients of interest (20), nine out of 20 (45%) were diagnosed with invasive carcinoma, one out of 20 (5%) with ductal carcinoma in situ, and 10 out of 20 (50%) with the same benign diagnosis (2 patients had excisional biopsies while 8 patients had a repeat ultrasound needle core biopsy). The remaining 16 cases were classified as benign following a comprehensive multidisciplinary team (MDT) discussion, taking into account the collective clinical, pathological, and radiological findings. Hence, the analysis included a total of 26 patients. Overall, 26 out of the 828 (3.14%) patients were identified as having discordant radiological findings.

Histopathological biopsy revealed that 6/26 (23.07%) patients were diagnosed with normal breast tissue, 8/26 (30.76%) with fibrocystic changes, 7/26 (26.92%) with inflammatory lesions (mastitis/granulomatous inflammation), 3/26 (11.53%) with fibroepithelial lesions, 1/26 (3.84%) with sclerosing adenosis, and 1/26 (3.84%) with atypical ductal hyperplasia (Fig. 2).

Fig. 2

Histopathological biopsy results and frequency. Column chart delineating the histopathological diagnoses and their frequency. The most prevalent diagnosis is fibrocystic changes, followed by inflammatory lesions (26.92%)

The analysed study population (with definite benign pathology) was all female, with 24/26 (92%) being 40 years of age or older. While 40/50 (80%) of the control group were 40 years of age or older.

Regarding the mass location, 15 lesions (57.69%) were located on the right side. The radiological assessment encompassed the examination of factors such as the side affected, posterior features, echogenicity, margins, shape, and size.

Univariable analysis

Chi-square analysis was employed to analyse each variable category between cases and controls; the results are presented in Table 1. The two groups differed significantly in terms of posterior features (p < 0.001), margins (p < 0.001), echogenicity (p = 0.002), and shape (p = 0.022), but not in terms of size or laterality. The posterior features observed were as follows: 8/26 (30.8%) patients showed shadowing, 6/26 (23.1%) patients showed enhancement, and 12/26 (46.2%) patients showed no abnormalities. The margin status was characterised as confined in 14/26 patients (53.8%), spiculated in 8/26 patients (30.8%), indistinct in 3/26 patients (11.5%), and angular in 1/26 patients (3.8%). Furthermore, 18/26 (69.2%) patients exhibited hypoechoic characteristics, 15/26 (57.7%) had irregular shape, and 15/26 (57.7%) had radiological diameters ranging from 1 cm to 3 cm. These results suggested potential differences in these radiological characteristics between malignant and benign breast masses.

In the control group of malignant breast lesions classified as BIRADS IVc, 31/50 (62%) were located on the right side, and the majority of cases exhibited the following characteristics: out of the total (50 controls), 43 cases (86%) exhibited shadowing, 50 cases (100%) were characterised as hypoechoic, 22 cases (44%) had spiculated borders, 43 cases (82%) had irregular shapes, and 72% had a radiological size between 1 cm and 3 cm.

Table 1 Chi-Square analysis of Ultrasound descriptors in breast masses (cases vs. controls)

Multivariable logistic regression analysis

A binary logistic regression analysis was conducted using IBM SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA) to examine the relationship between BIRADS IVc/V lesions and their biopsy-based diagnosis while adjusting for potential confounding factors. The logistic regression results, including odds ratios (OR), confidence intervals (CI), and p-values, are presented in Table 2. The results indicated that the previously significant variables (margins and echogenicity) did not retain statistical significance when assessed independently within the multivariable model. Shape analysis revealed a substantial increase in lesions with a benign diagnosis when they had an “oval shape,” with a confidence interval of 4.536 (1.114–18.478) (p = 0.035). The posterior shadowing showed a significant decrease with a confidence interval of 0.303 (0.091–1.002), p-value = 0.050. Nevertheless, there was no noteworthy correlation observed with individuals aged over 40 years, tumours larger than 3.0 cm in size, or tumours with a spiculated margin. Furthermore, there was evidence of right-sided laterality, with a confidence interval of 0.468 (0.128–1.708). The loss of significance in the logistic regression analysis might be contributed to different factors, including: the associations observed in the chi-square analysis might have been confounded by other variables such as age, size, or laterality that could have interacted with the radiological descriptors to influence the outcome. Also, logistic regression requires a larger sample size to detect significant associations because it adjusts for multiple covariates simultaneously.

In 83.32% of the research cases and 92% of the control cases, the biopsy size collected for histopathology was greater than 2.0 cm.

Table 2 Logistic regression analysis of Radiological variables for breast masses

Discussion

Descriptions related to radiology

This study’s findings offer useful insights into the diagnosis and features of benign breast lesions categorised as BIRADS IVc or V within a single medical facility from 2015 to 2022. Of the 828 tumours that were found, a substantial majority (94.6%) were initially classified as malignant based on the initial biopsy. This underscores the significance of conducting additional assessments for these patients. Nevertheless, it is important to mention that 5.31% of cases were subsequently determined to be benign following histological testing. This highlights the difficulties and intricacies involved in precisely detecting breast lesions and emphasises the significance of employing precise diagnostic techniques to prevent unwanted procedures and distress for patients.

The BIRADS radiological method is extensively employed to categorise breast lesions and aid radiologists in the pursuit of biopsies for worrisome lesions. Nevertheless, certain radiological characteristics can impede the accurate detection of both noncancerous and cancerous abnormalities. While it is important for radiologists to minimise the use of unneeded biopsies, there are situations when it might be challenging to determine if a biopsy is truly needed. Chung et al. discovered that certain noncancerous breast conditions, such as fibrocystic alterations and PASH (pseudoangiomatous stromal hyperplasia), frequently exhibit worrisome imaging characteristics that necessitate biopsy for confirmation and may even necessitate subsequent biopsies for further evaluation [11].

The categorisation of each group is contingent upon the imaging descriptors of the breast mass, including breast composition, density, calcification, asymmetry, location, and other characteristics. Among these characteristics, a tumour that has a high density, an irregular shape, and spiculated edges is most likely to be malignant [5]. This study revealed a substantial increase in patients with BIRADS IVc/V masses and a benign diagnosis on biopsy of lesions that had an “oval shape,” p = 0.035. The posterior shadowing exhibited a considerable decline (p = 0.050). No significant associations were detected with regard to age over 40 years (CI 3.850 (0.385–38.531)), radiological large size over 3.0 cm (CI 2.379 (0.701–8.080)), spiculated margin (CI 1.127 (0.651–1.951)), or right-sided laterality (CI 0.468 (0.128–1.708)).

Previous research has shown that the positive predictive values (PPVs) for malignancy in subcategories BIRADS IVc varies, for example 74% in one study [12] and 97% in another study [13]. This indicates that while BI-RADS IVc lesions have a higher PPV for malignancy, a significant proportion are benign. Advanced imaging modalities, such as elastography [14], diffusion-weighted MRI [15], or artificial intelligence-based image analysis [16], may be utilised to better distinguish between benign and malignant BIRADS IVc lesions, hence minimising the need for repeat biopsy.

Diagnosis of pathological conditions

BIRADS IVc/V lesions strongly indicate the presence of cancer. Nevertheless, a portion of these lesions are classified as benign. A study revealed that the majority of BI-RADS IV lesions exhibited fibrocystic alterations upon histological evaluation, with ductal carcinoma in situ (DCIS) being the second most prevalent finding [6]. The primary objective of the BIRADS system is to establish a uniform method for assessing the likelihood of breast lesions being cancerous. Previous studies conducted by Baker et al. and Berg et al. have highlighted the significant variability in lesion description and therapy [7, 8].

Among our study patients, of the total number of patients with BIRADS IVc/V disease, 8/26 (30.76%) had fibrocystic alterations. Noticeably, 30.76% of patients exhibited spiculated breast masses that were identified as potentially cancerous on radiography but were ultimately confirmed as benign conditions such as sclerosing adenosis, granulomatous mastitis, and fibrocystic alterations by histological examination. In their study, Ji Hyun Youk et al. examined cases with benign lesions that were inconsistent with each other and emphasised the significance of prompt communication between radiologists and pathologists in determining if a repeat biopsy is necessary [17].

Ductal carcinoma in situ might manifest as a mass lesion, often accompanied by calcifications and persistent enhancement, as described by Yamada [18]. The case of ductal carcinoma in situ in our study, which was not detected in the initial biopsy, was presented as an asymmetrical angular mass lesion lacking calcification or enhancement.

It is important to highlight the demographic and anatomical aspects of the study population. The majority of patients were above the age of 40, consistent with the higher incidence of breast cancer in older adults. The majority of the lesions were located on the right side and in the upper outer quadrant, which corresponds to the established distribution patterns of breast tumors. Furthermore, the coexistence of both mass and non-mass lesions in patients highlights the varied characteristics of breast lesions and emphasises the necessity for customised diagnostic and therapeutic strategies.

Eighteen of the 26 patients were followed up for at least two years; all had stable lesions with no concerning features on mammogram. One patient underwent bilateral breast reduction surgery, and one developed nodal lymphoma. Eight patients were lost to follow-up.

Size of the biopsy

An optimal biopsy size is crucial for achieving a precise diagnosis and preventing the oversight of serious illnesses. Hyun Kyung Jung discovered that performing an ultrasound-guided core biopsy on breast lesions measuring 2.0 cm or larger resulted in a 98.6% accuracy [19]. Our investigation identified ten instances of malignancy that were not detected during the initial biopsy. Among these cases, three (30%) involved little tissue samples measuring less than 1.0 cm, which was insufficient for accurate evaluation. Consequently, a more comprehensive biopsy is necessary to obtain a representative sample. In the Selvi Radhakrishna study, of the total number of patients, five individuals (1.07%) had a core biopsy that did not provide a conclusive diagnosis. As a result, these patients need to undergo an excision biopsy to obtain a definitive diagnosis [20]. Holzer-Fruehwald et al. highlighted that there is no precise cutoff value for tumour size that is correlated with malignancy [21]. Nevertheless, she discovered that tumours larger than 10.0 mm exhibited a malignancy risk of 32.5%. The average size of malignant tumours in our study exceeded 10.0 mm. However, five patients were benign and had diameters greater than 30.0 mm, three of whom were diagnosed with granulomatous mastitis.

Limitations

This study was limited by its small sample size. This study was a retrospective cross-sectional analysis in which the training level of the radiologists during primary radiological evaluation was not established. Nevertheless, the data were gathered in a sizable tertiary referral hospital encompassing a vast area in Jordan. Previous research has indicated that the level of expertise possessed by radiologists has a minimal impact on positive predictive values (PPVs) for subtypes IVa-IVc [9].

Conclusion

The most prevalent histopathological diagnosis in our study for BIRADS IVc/V patients with benign breast lesions was fibrocystic change. From a radiological perspective, only the oval shape of the lesion correlates with a benign diagnosis. Conversely, the presence of posterior shadowing was significantly less common. Additional radiological descriptors were unable to distinguish between benign and malignant breast lesions. This study encourages multidisciplinary decision-making among radiologists, pathologists, and clinicians to determine if a repeat biopsy is warranted in such cases.

The results of this study emphasise the importance of continuous research to improve breast lesions diagnosis and treatment. It suggests reducing false-positive rates and unnecessary biopsies in benign lesions and establishing clinical guidelines for repeating biopsy in patients with high-risk radiological features. The study also suggests exploring other diagnostic modalities like sonoelastography and incorporating deep learning and artificial intelligence in decision-making to eliminate unnecessary procedures.