Biomineralogical signatures of breast microcalcifications

Optimizing breast lesions diagnosis and decision-making with a deep learning fusion model integrating ultrasound and mammography: a dual-center retrospective study

BACKGROUND

This study aimed to develop a BI-RADS network (DL-UM) via integrating ultrasound (US) and mammography (MG) images and explore its performance in improving breast lesion diagnosis and management when collaborating with radiologists, particularly in cases with discordant US and MG Breast Imaging Reporting and Data System (BI-RADS) classifications.

METHODS

We retrospectively collected image data from 1283 women with breast lesions who underwent both US and MG within one month at two medical centres and categorised them into concordant and discordant BI-RADS classification subgroups. We developed a DL-UM network via integrating US and MG images, and DL networks using US (DL-U) or MG (DL-M) alone, respectively. The performance of DL-UM network for breast lesion diagnosis was evaluated using ROC curves and compared to DL-U and DL-M networks in the external testing dataset. The diagnostic performance of radiologists with different levels of experience under the assistance of DL-UM network was also evaluated.

RESULTS

In the external testing dataset, DL-UM outperformed DL-M in sensitivity (0.962 vs. 0.833, P = 0.016) and DL-U in specificity (0.667 vs. 0.526, P = 0.030), respectively. In the discordant BI-RADS classification subgroup, DL-UM achieved an AUC of 0.910. The diagnostic performance of four radiologists improved when collaborating with the DL-UM network, with AUCs increased from 0.674-0.772 to 0.889-0.910, specificities from 52.1%-75.0 to 81.3-87.5% and reducing unnecessary biopsies by 16.1%-24.6%, particularly for junior radiologists. Meanwhile, DL-UM outputs and heatmaps enhanced radiologists' trust and improved interobserver agreement between US and MG, with weighted kappa increased from 0.048 to 0.713 (P < 0.05).

CONCLUSIONS

The DL-UM network, integrating complementary US and MG features, assisted radiologists in improving breast lesion diagnosis and management, potentially reducing unnecessary biopsies.

Improving Prognostic Value in Invasive Triple Negative Breast Cancer Through a Combined Nomogram Approach

OBJECTIVES

To investigate the potential prognostic value of ultrasound (US) features in conjunction with pathological markers and to develop a preliminary working model for predicting poor outcomes in patients with invasive triple-negative breast cancer (TNBC).

METHODS

From January 2012 to December 2018, we enrolled 209 TNBC patients treated with standard therapy, systematically gathered data on US parameters, stromal tumor-infiltrating lymphocytes (TILs), lymphovascular invasion (LVI) status, and other relevant information, and recorded follow-up data. A nomogram combining AJCC staging with US score, stromal TILs, and LVI was constructed and validated to predict poor outcomes, defined as recurrence or death, in patients with invasive TNBC.

RESULTS

The US score of 4 was best related to poor outcomes in patients with TNBC (HR 3.87, P = .015). In the training set, the nomogram had a considerably greater prognostic value [area under the curve (AUC), 0.74 vs. 0.64, P = .045] than AJCC staging alone, and it was comparable to that of the validation set (AUC, 0.71 vs. 0.63, P = .804). An acceptable consistency between the nomogram-predicted and actual survival probabilities was found both in the training and validation sets, with Brier scores of 0.15 and 0.13, respectively.

CONCLUSIONS

The incorporation of AJCC stage with US score, stromal TILs, and LVI improved the model performance for predicting poor outcomes in patients with invasive TNBC compared to routine AJCC staging alone.

Implications of Mineralization Biomarkers in Breast Cancer Outcomes Beyond Calcifications

The aim of this work was to explore the biomarkers associated with epithelial to mesenchymal transition (EMT) and mineralization processes as new prognostic factors across different breast cancer phenotypes. To this end, 133 breast biopsies, including benign and malignant lesions, with or without microcalcifications, were retrospectively collected. Immunohistochemical analysis was performed to evaluate the expression of vimentin, BMP-2, BMP-4, RANKL, Runx2, OPN, PTX3, and SDF-1, while Kaplan-Meier plots were used to assess their prognostic impact on overall survival in a dataset of 2976 breast cancer patients. The expression of vimentin, BMP-2, BMP-4, and SDF-1 was significantly higher in malignant lesions compared to benign ones, regardless of the presence of microcalcifications. Notably, these markers showed no correlation with traditional prognostic factors, such as tumor grade or hormone receptor status. The bioinformatics analysis provided valuable insights into the possible prognostic and therapeutic significance of BMP-2, BMP-4, SDF-1, and vimentin in breast cancer. In fact, all these biomarkers impact on the overall survival in specific molecular breast cancer types. In addition, high expression of SDF-1 and vimentin is able to predict the response to chemotherapy. The findings here reported suggest that vimentin, BMP-2, BMP-4, and SDF-1 could be independent prognostic biomarkers in breast cancer, providing insights beyond traditional clinical factors.

Characterization of dolomite and calcite microcalcifications in human breast tissue

Pathological crystallization within soft tissues often yields biominerals with properties differing from those of their geological or synthetic counterparts. Microcalcifications (MCs) are abundant in breast tumors, particularly in non-invasive lesions, such as ductal carcinoma in situ (DCIS). Given the challenge of predicting DCIS progression into invasive cancer, it has been suggested that MCs can be leveraged to inform DCIS prognosis. The predominant type of breast MCs are those containing calcium phosphates (CaP), whose crystal properties are commonly held to correlate with malignancy. Less common are non-CaP minerals, which have received less attention, as they are associated mainly with benign lesions. Here, we conducted a retrospective study of tissue samples collected from patients who were originally diagnosed with DCIS and whose current medical status is known. We examined the elemental composition, morphology, and crystal phases of 398 MCs, aiming to investigate potential correlations between MC crystal properties and the progression of DCIS. Our findings revealed primarily non-CaP MCs, an observation that was likely made possible only by the tissue processing methodology employed, which did not involve harsh conditions. We found that non-CaP MCs were abundant in DCIS lesions, that they exhibited diverse morphologies and sizes, and that they were composed of calcite and dolomite. Dolomite formation in cancer has not been reported previously and may be linked to pH fluctuations in the tumor microenvironment. The small size of DCIS lesions often requires pathologists to use the entire sample, thus reducing the number of samples available for further research. Nonetheless, despite our limited sample size, the observed trend indicated an association of dolomite MCs with DCIS lesions that progressed into invasive cancer over time.

Quality-Assured Analysis of PIK3CA Mutations in Hormone Receptor-Positive/Human Epidermal Growth Factor Receptor 2-Negative Breast Cancer Tissue

In precision oncology, reliable testing of predictive molecular biomarkers is a prerequisite for optimal patient treatment. Interlaboratory comparisons are a crucial tool to verify diagnostic performance and reproducibility of one's approach. Herein is described the design and results of the first recurrent, internationally performed PIK3CA (phosphatidylinositol-4,5-bisphosphate 3 kinase catalytic subunit α) breast cancer tissue external quality assessment (EQA), organized by German Quality in Pathology GmbH and started in 2021. After the internal pretesting phase performed by the (lead) panel institutes, in both 2021 and 2022, each EQA test set comprised n = 10 tissue samples of hormone receptor-positive, human epidermal growth factor receptor 2-negative invasive breast cancer that had to be analyzed and reported by the participants. In 2021, the results were evaluated separately for German-speaking countries (part 1) and international laboratories (part 2). In 2022, the EQA was performed across the European Union. The EQA success rates were 84.6% (n = 11/13), 88.6% (n = 39/44), and 87.9% (n = 29/33) for EQA 2021 part 1, part 2, and EQA 2022, respectively. The most commonly used methods were next-generation sequencing and mutation-/allele-specific qualitative PCR-based assays. In summary, this recurrent PIK3CA EQA proved to be a suitable approach to obtain an international overview of methods used for PIK3CA mutation analysis, to evaluate them qualitatively, and identify the strengths and weaknesses of individual methods.

The interplay between crystallinity and the levels of Zn and carbonate in synthetic microcalcifications directs thyroid cell malignancy

One of the key challenges in diagnosing thyroid cancer lies in the substantial percentage of indeterminate diagnoses of thyroid nodules that have undergone ultrasound-guided fine-needle aspiration (FNA) biopsy for cytological evaluation. This delays the definitive diagnosis and treatment plans. We recently demonstrated that hydroxyapatite microcalcifications (MCs) aspirated from thyroid nodules may aid nodule diagnosis based on their composition. In particular, Zn-enriched MCs have emerged as potential cancer biomarkers. However, a pertinent question remains: is the elevated Zn content within MCs a consequence of cancer, or do the Zn-enriched MCs encourage tumorigenesis? To address this, we treated the human thyroid cancer cell line MDA-T32 with synthetic MC analogs comprising hydroxyapatite crystals with varied pathologically relevant Zn fractions and assessed the cellular response. The MC analogs exhibited an irregular surface morphology similar to FNA MCs observed in cancerous thyroid nodules. These MC analogs displayed an inverse relationship between Zn fraction and crystallinity, as shown by X-ray diffractometry. The zeta potential of the non-Zn-bearing hydroxyapatite crystals was negative, which decreased once Zn was incorporated into the crystal. The MC analogs were not cytotoxic. The cellular response to exposure to these crystals was evaluated in terms of cell migration, proliferation, the tendency of the cells to form multicellular spheroids, and the expression of cancer markers. Our findings suggest that, if thyroid MCs play a role in promoting cancerous behavior in vivo, it is likely a result of the interplay of crystallinity with Zn and carbonate fractions in MCs.

Translating microcalcification biomarker information into the laboratory: A preliminary assessment utilizing core biopsies obtained from sites of mammographic calcification

Rationale and objectives

The potential of breast microcalcification chemistry to provide clinically valuable intelligence is being increasingly studied. However, acquisition of crystallographic details has, to date, been limited to high brightness, synchrotron radiation sources. This study, for the first time, evaluates a laboratory-based system that interrogates histological sections containing microcalcifications. The principal objective was to determine the measurement precision of the laboratory system and assess whether this was sufficient to provide potentially clinical valuable information.

Materials and methods

Sections from 5 histological specimens from breast core biopsies obtained to evaluate mammographic calcification were examined using a synchrotron source and a laboratory-based instrument. The samples were chosen to represent a significant proportion of the known breast tissue, mineralogical landscape. Data were subsequently analysed using conventional methods and microcalcification characteristics such as crystallographic phase, chemical deviation from ideal stoichiometry and microstructure were determined.

Results

The crystallographic phase of each microcalcification (e.g., hydroxyapatite, whitlockite) was easily determined from the laboratory derived data even when a mixed phase was apparent. Lattice parameter values from the laboratory experiments agreed well with the corresponding synchrotron values and, critically, were determined to precisions that were significantly greater than required for potential clinical exploitation.

Conclusion

It has been shown that crystallographic characteristics of microcalcifications can be determined in the laboratory with sufficient precision to have potential clinical value. The work will thus enable exploitation acceleration of these latent microcalcification features as current dependence upon access to limited synchrotron resources is minimized.

Calcium-based biomaterials: Unveiling features and expanding applications in osteosarcoma treatment

Calcium, an indispensable element in bone tissues, plays a crucial role in various cellular processes involved in cancer progression. Its ubiquitous yet spatially distinct distribution in the body presents an opportunity to target calcium homeostasis as a novel strategies for cancer treatment, with specific advantages in osteosarcoma therapy. In this comprehensive review, we retrospect the calcium biology intersected with cancer progression, highlight the unveiling features of calcium-based biomaterials in regulating both bone homeostasis and cancer development. We also provide an overview of recent breakthroughs in cancer therapy that leverage calcium biomaterials, showcasing their potential to serve as versatile, customizable platforms for osteosarcoma treatment and as reservoirs for supporting bone reconstruction.

A Mammography-Based Radiomic Nomogram for Predicting Malignancy in Breast Suspicious Microcalcifications

RATIONALE AND OBJECTIVES

Preoperative accurate identification of benign and malignant breast lesions is vital for patients to achieve individualized treatment. This study aimed to develop and validate a mammography-based radiomic nomogram for predicting malignant risk of breast suspicious microcalcifications (MCs).

MATERIALS AND METHODS

496 patients with histologically confirmed breast suspicious MCs were randomly divided into the training set (n = 346) and validation set (n = 150). Radiomics features was extracted from the craniocaudal and mediolateral oblique images. Least absolute shrinkage and selection operator algorithm were used to select radiomics features, then radiomics score (Rad-score) was calculated. Univariate analysis was used to identify malignant MCs-related clinical independent risk factors. Multivariate logistic regression was used to establish a clinical-radiomics model by incorporating Rad-score and clinic factors. A nomogram was developed to visualize the clinical-radiomics model. The receiver operating characteristic curve, calibration curve and decision curve analysis (DCA) were used to evaluate the performance of the nomogram.

RESULTS

The Rad-score was consisted of 29 optimal radiomics features. We developed a nomogram by incorporating Rad-score, menopause status, MCs morphology and distribution, the area under the curve value of the combined model was 0.926(95% confidence interval [CI]: 0.878-0.975) for the validation set. The calibration curves and DCA indicated the combined model had favorable calibration and clinical utility.

CONCLUSION

The combined model could be considered as a potential imaging marker to predict malignant risk of breast suspicious MCs.

Multicellular spheroids containing synthetic mineral particles: an advanced 3D tumor model system to investigate breast precancer malignancy potential according to the mineral type

Mineral particles that form in soft tissues in association with disease conditions are heterogeneous in their composition and physiochemical properties. Hence, it is challenging to study the effect of mineral type on disease progression in a high-throughput and realistic manner. For example, most early breast precancer lesions, termed ductal carcinoma in situ (DCIS), contain microcalcifications (MCs), calcium-containing pathological minerals. The most common type of MCs is calcium phosphate crystals, mainly carbonated apatite; it is associated with either benign or malignant lesions. In vitro studies indicate that the crystal properties of apatite MCs can affect breast cancer progression. A less common type of MCs is calcium oxalate dihydrate (COD), which is almost always found in benign lesions. We developed a 3D tumor model of multicellular spheroids of human precancer cells containing synthetic MC analogs that link the crystal properties of MCs with the progression of breast precancer to invasive cancer. Using this 3D model, we show that apatite crystals induce Her2 overexpression in DCIS cells. This tumor-triggering effect is increased when the carbonate fraction in the MCs decreases. COD crystals, in contrast, decrease Her2 expression in the spheroids, even compared with a control group with no added MC analogs. Furthermore, COD decreases cell proliferation and migration in DCIS monolayers compared to untreated cells and cells incubated with apatite crystals. This finding suggests that COD is not randomly located only in benign lesions-it may actively contribute to suppressing precancer progression in its surroundings. Our model provides an easy-to-manipulate platform to better understand the interactions between mineral particles and their biological microenvironment. A better understanding of the effect of the crystal properties of MCs on precancer progression will potentially provide new directions for better precancer prognosis and treatment.