survex: an R package for explaining machine learning survival models.

SUMMARY: Due to their flexibility and superior performance, machine learning models frequently complement and outperform traditional statistical survival models. However, their widespread adoption is hindered by a lack of user-friendly tools to explain their internal operations and prediction rationales. To tackle this issue, we introduce the survex R package, which provides a cohesive framework for explaining any survival model by applying explainable artificial intelligence techniques. The capabilities of the proposed software encompass understanding and diagnosing survival models, which can lead to their improvement. By revealing insights into the decision-making process, such as variable effects and importances, survex enables the assessment of model reliability and the detection of biases. Thus, transparency and responsibility may be promoted in sensitive areas, such as biomedical research and healthcare applications. AVAILABILITY AND IMPLEMENTATION: survex is available under the GPL3 public license at https://github.com/modeloriented/survex and on CRAN with documentation available at https://modeloriented.github.io/survex.

An Explainable Deep Learning Classifier of Bovine Mastitis Based on Whole-Genome Sequence Data-Circumventing the p >> n Problem.

The serious drawback underlying the biological annotation of whole-genome sequence data is the p >> n problem, which means that the number of polymorphic variants (p) is much larger than the number of available phenotypic records (n). We propose a way to circumvent the problem by combining a LASSO logistic regression with deep learning to classify cows as susceptible or resistant to mastitis, based on single nucleotide polymorphism (SNP) genotypes. Among several architectures, the one with 204,642 SNPs was selected as the best. This architecture was composed of two layers with, respectively, 7 and 46 units per layer implementing respective drop-out rates of 0.210 and 0.358. The classification of the test data resulted in AUC = 0.750, accuracy = 0.650, sensitivity = 0.600, and specificity = 0.700. Significant SNPs were selected based on the SHapley Additive exPlanation (SHAP). As a final result, one GO term related to the biological process and thirteen GO terms related to molecular function were significantly enriched in the gene set that corresponded to the significant SNPs. Our findings revealed that the optimal approach can correctly predict susceptibility or resistance status for approximately 65% of cows. Genes marked by the most significant SNPs are related to the immune response and protein synthesis.

SARS-CoV-2 Infection in Patients Treated with Azacitidine and Venetoclax for Acute Leukemia: A Report of a Case Series Treated in a Single Institution.

INTRODUCTION: Venetoclax combined with azacitidine (AZA-VEN) constitutes an option for the treatment of acute myeloid leukemia. There are, however, no data on the COVID-19 incidence and outcome in patients treated with AZA-VEN. METHODS: Patients with acute leukemia treated with AZA-VEN at a single institution were included in this prospective observational study. RESULTS: Thirteen patients were enrolled, 46% with treatment-naïve, and 56% with relapsed/refractory disease. Fifty-four percent of patients were males; the median age was 69 years. Six patients (46%) developed COVID-19 during the observation time. The median time to COVID-19 was 24 days from the initiation of AZA-VEN. The 2-month cumulative incidence of COVID-19 was 46.2%. Two patients (33%) succumbed to COVID-19. The 100-day COVID-19-free survival from AZA-VEN initiation was 61%. The median follow-up time was 4.3 months. DISCUSSION/CONCLUSION: COVID-19 constitutes a frequent complication of AZA-VEN treatment in the era of the COVID-19 pandemic, leading to death in a significant proportion of patients.

miRNA-Seq Tissue Diagnostic Signature: A Novel Model for NSCLC Subtyping.

Non-small cell lung cancer (NSCLC) encompasses distinct histopathological subtypes, namely adenocarcinoma (AC) and squamous cell lung carcinoma (SCC), which require precise differentiation for effective treatment strategies. In this study, we present a novel molecular diagnostic model that integrates tissue-specific expression profiles of microRNAs (miRNAs) obtained through next-generation sequencing (NGS) to discriminate between AC and SCC subtypes of NSCLC. This approach offers a more comprehensive and precise molecular characterization compared to conventional methods such as histopathology or immunohistochemistry. Firstly, we identified 31 miRNAs with significant differential expression between AC and SCC cases. Subsequently, we constructed a 17-miRNA signature through rigorous multistep analyses, including LASSO/elastic net regression. The signature includes both upregulated miRNAs (hsa-miR-326, hsa-miR-450a-5p, hsa-miR-1287-5p, hsa-miR-556-5p, hsa-miR-542-3p, hsa-miR-30b-5p, hsa-miR-4728-3p, hsa-miR-450a-1-3p, hsa-miR-375, hsa-miR-147b, hsa-miR-7705, and hsa-miR-653-3p) and downregulated miRNAs (hsa-miR-944, hsa-miR-205-5p, hsa-miR-205-3p, hsa-miR-149-5p, and hsa-miR-6510-3p). To assess the discriminative capability of the 17-miRNA signature, we performed receiver operating characteristic (ROC) curve analysis, which demonstrated an impressive area under the curve (AUC) value of 0.994. Our findings highlight the exceptional diagnostic performance of the miRNA signature as a stratifying biomarker for distinguishing between AC and SCC subtypes in lung cancer. The developed molecular diagnostic model holds promise for providing a more accurate and comprehensive molecular characterization of NSCLC, thereby guiding personalized treatment decisions and improving clinical management and prognosis for patients.

miRNA Studies in Glaucoma: A Comprehensive Review of Current Knowledge and Future Perspectives.

Glaucoma, a neurodegenerative disorder that leads to irreversible blindness, remains a challenge because of its complex nature. MicroRNAs (miRNAs) are crucial regulators of gene expression and are associated with glaucoma and other diseases. We aimed to review and discuss the advantages and disadvantages of miRNA-focused molecular studies in glaucoma through discussing their potential as biomarkers for early detection and diagnosis; offering insights into molecular pathways and mechanisms; and discussing their potential utility with respect to personalized medicine, their therapeutic potential, and non-invasive monitoring. Limitations, such as variability, small sample sizes, sample specificity, and limited accessibility to ocular tissues, are also addressed, underscoring the need for robust protocols and collaboration. Reproducibility and validation are crucial to establish the credibility of miRNA research findings, and the integration of bioinformatics tools for miRNA database creation is a valuable component of a comprehensive approach to investigate miRNA aberrations in patients with glaucoma. Overall, miRNA research in glaucoma has provided significant insights into the molecular mechanisms of the disease, offering potential biomarkers, diagnostic tools, and therapeutic targets. However, addressing challenges such as variability and limited tissue accessibility is essential, and further investigations and validation will contribute to a deeper understanding of the functional significance of miRNAs in glaucoma.

KRAB ZNF explorer-the online tool for the exploration of the transcriptomic profiles of KRAB-ZNF factors in The Cancer Genome Atlas.

SUMMARY: Krüppel-associated box domain zinc finger protein (KRAB ZNF) explorer is a web-based tool for comprehensive characterization of the mRNA expression status of KRAB-ZNF transcription factors in the data from The Cancer Genome Atlas study. Key functionalities cover a comparative analysis of KRAB-ZNF expression between normal and cancer tissues, an association of KRAB-ZNF expression with various pathological features, including survival analysis, association with global DNA methylation status, analyses of KRAB-ZNF isoform expressions and analysis of KRAB-ZNF levels in normal tissues. AVAILABILITY AND IMPLEMENTATION: KRAB ZNF explorer is available at http://mi2.mini.pw.edu.pl: 8080/KRAB_ZNF/. The source code for shiny application is available at: https://github.com/MI2DataLab/KRAB_ZNF/tree/master/app and the source code for analyses and precalculations are available at: https://github.com/MI2DataLab/KRAB_ZNF/tree/master/work.

Azacitidine for relapse of acute myeloid leukemia or myelodysplastic syndrome after allogeneic hematopoietic stem cell transplantation, multicenter PALG analysis.

OBJECTIVES: Relapse of myelodysplastic syndromes (MDS) or acute myeloid leukemia (AML) after allogeneic hematopoietic stem cell transplantation (allo-HSCT) belongs to the major causes of treatment failure. METHODS: Retrospective multicenter analysis of patients diagnosed with AML or MDS who had hematological relapse after allo-HSCT and were treated with azacitidine for this indication. RESULTS: Twenty-three patients receiving azacitidine as the first treatment of relapse (Group_1) and 8 patients receiving azacitidine after other treatment of relapse (Group_2) were included. There were 68% males, median age at initiation of azacitidine was 53 years (15-66). Median time to relapse was 3.5 months and 6.3 months in Group_1 and Group_2, respectively; median time from relapse to azacitidine 0.2 and 2.3 months. Azacitidine 75 mg/m2 , days 1-7, was administered in 78% and 75% of patients in Group_1 and Group_2, concomitant DLI in 48% and 50%. With median follow-up of 4.7 and 13.6 months, the median overall survival was 5.9 and 9.5 months. 17% and 37.5% patients proceeded to salvage allo-HSCT, with median OS of 11.6 months and not reached respectively. CONCLUSIONS: Azacitidine treatment for hematological relapse is associated with poor outcome; nevertheless, a proportion of patients may benefit from it, including patients receiving subsequent salvage allo-HSCT.

Checklist for responsible deep learning modeling of medical images based on COVID-19 detection studies.

The sudden outbreak and uncontrolled spread of COVID-19 disease is one of the most important global problems today. In a short period of time, it has led to the development of many deep neural network models for COVID-19 detection with modules for explainability. In this work, we carry out a systematic analysis of various aspects of proposed models. Our analysis revealed numerous mistakes made at different stages of data acquisition, model development, and explanation construction. In this work, we overview the approaches proposed in the surveyed Machine Learning articles and indicate typical errors emerging from the lack of deep understanding of the radiography domain. We present the perspective of both: experts in the field - radiologists and deep learning engineers dealing with model explanations. The final result is a proposed checklist with the minimum conditions to be met by a reliable COVID-19 diagnostic model.

Prognostic Role of Tumoral PD-L1 and IDO1 Expression, and Intratumoral CD8+ and FoxP3+ Lymphocyte Infiltrates in 132 Primary Cutaneous Merkel Cell Carcinomas.

The association of immune markers and clinicopathologic features and patient outcome has not been extensively studied in Merkel cell carcinoma (MCC). We correlated tumoral PD-L1 and IDO1 expression, and intratumoral CD8+ and FoxP3+ lymphocytes count with clinicopathologic variables, Merkel cell polyomavirus (MCPyV) status, and patient outcomes in a series of 132 MCC. By univariate analyses, tumoral PD-L1 expression >1% and combined tumoral PD-L1 >1% and high intratumoral FoxP3+ lymphocyte count correlated with improved overall survival (OS) (p = 0.016, 0.0072), MCC-specific survival (MSS) (p = 0.019, 0.017), and progression-free survival (PFS) (p = 0.043, 0.004, respectively). High intratumoral CD8+ and FoxP3+ lymphocyte count correlated with longer MSS (p = 0.036) and improved PFS (p = 0.047), respectively. Ulceration correlated with worse OS and worse MSS. Age, male gender, and higher stage (3 and 4) significantly correlated with worse survival. MCPyV positivity correlated with immune response. By multivariate analyses, only ulceration and age remained as independent predictors of worse OS; gender and stage remained for shorter PFS. Tumoral PD-L1 expression and increased density of intratumoral CD8+ lymphocytes and FoxP+ lymphocytes may represent favorable prognosticators in a subset of MCCs. Tumoral PD-L1 expression correlated with intratumoral CD8+ and FoxP3+ lymphocytes, which is supportive of an adaptive immune response.

Impact of pre-transplantation sinus abnormalities assessed by Lund-Mackay scoring system on sinusitis occurrence after allogeneic hematopoietic cell transplantation in acute leukemia patients.

INTRODUCTION: Screening sinonasal evaluation is routinely performed before allogeneic hematopoietic cell transplantation (allo-HCT), however, data supporting such evaluation is inconsistent. O b j e c t i v e s: Assessment of the utility of screening sinonasal evaluation with computed tomography (CT). METHODS: A retrospective analysis of acute leukemia patients who underwent allo-HCT, for whom screening sinonasal CT scans were reevaluated, and for whom Lund-Mackay score (LMS) was calculated. R e s u l t s: Forty-eight patients, the median age at allo-HCT 38 years (18-58), 52% males, were included. 79% had acute myeloid leukemia (AML), 21% acute lymphoblastic leukemia (ALL). Conditioning intensity was myeloablative in 96% of patients, 21% of patients received total body irradiation. 19% of patients had a history of sinusitis before allo-HCT. Screening sinus CT was performed a median of 22 days before allo-HCT. The median LMS was 1 point (0- 10). The severity of sinus abnormalities was: no abnormalities (31%), mild (67%), moderate (2%), severe (0%). Mucosal thickening was the most frequent abnormality (69%). Eleven patients experienced sinusitis after a median of 93 days (11-607) after allo-HCT. 1-year cumulative incidence of sinusitis was 22%. No threshold of LMS and no type of sinus abnormalities were correlated with sinusitis development after allo-HCT. Mild sinus disease at screening did not negatively impact survival in comparison to no sinus disease. C o n c l u s i o n s: Despite the fact, that majority of analyzed patients had either no or mild sinus disease at screening a significant proportion of patients developed sinusitis after allo-HCT. Evaluation of LMS before allo-HCT did not help predict the development of sinusitis after the procedure.

Allogeneic Hematopoietic Stem Cell Transplantation for Paroxysmal Nocturnal Hemoglobinuria: Multicenter Analysis by the Polish Adult Leukemia Group.

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is the sole potential cure for paroxysmal nocturnal hemoglobinuria (PNH); however, the data on its utility in PNH are limited. This retrospective analysis of patients with PNH who underwent allo-HSCT in 11 Polish centers between 2002 and 2016 comprised 78 patients with PHN, including 27 with classic PNH (cPNH) and 51 with bone marrow failure-associated PNH (BMF/PNH). The cohort was 59% male, with a median age of 29 years (range, 12 to 65 years). There was a history of thrombosis in 12% and a history of hemolysis in 81%, and 92% required erythrocyte transfusions before undergoing allo-HSCT. No patient received eculizumab, and 26% received immunosuppressive treatment. The median time from diagnosis to allo-HSCT was 12 months (range, 1 to 127 months). Almost all patients (94%) received reduced-toxicity conditioning, 66% with treosulfan. The stem cell source was peripheral blood in 72% and an identical sibling donor in 24%. Engraftment occurred in 96% of the patients. With a median follow-up of 5.1 years in patients with cPNH and 3.2 years in patients with BMF/PNH, 3-year overall survival (OS) was 88.9% in the former and 85.1% in the latter (P = not significant [NS]). The 3-year OS for patients with/without thrombosis was 50%/92% (P = NS) in the cPNH group and 83.3%/85.3% (P = NS) in the BMF/PNH group. The 3-year OS for in the BMF/PNH patients with/without hemolysis was 93.9%/62.9% (hazard ratio, .13; P = .016). No other factors impacted OS. After allo-HSCT, the frequency of the PNH clone was reduced to 0%, <1%, and <2.4% in 48%, 48%, and 4% of cPNH patients and in 84%, 11%, and 5% of BMF/PNH patients, respectively. The frequency of acute graft-versus-host disease (GVHD) grade II-IV was 23%, and the cumulative 1-year incidence of extensive chronic GVHD was 10.8% in the BMF/PNH group and 3.7% in the cPNH group. Allo-HSCT is a valid option for PNH patients, effectively eliminating the PNH clone with satisfactory overall survival and acceptable toxicity. Reduced-toxicity conditioning with treosulfan is effective and safe in patients with cPNH and BMF/PNH.

High dimensional surrogacy: computational aspects of an upscaled analysis.

Identification of genomic biomarkers is an important area of research in the context of drug discovery experiments. These experiments typically consist of several high dimensional datasets that contain information about a set of drugs (compounds) under development. This type of data structure introduces the challenge of multi-source data integration. High-Performance Computing (HPC) has become an important tool for everyday research tasks. In the context of drug discovery, high dimensional multi-source data needs to be analyzed to identify the biological pathways related to the new set of drugs under development. In order to process all information contained in the datasets, HPC techniques are required. Even though R packages for parallel computing are available, they are not optimized for a specific setting and data structure. In this article, we propose a new framework, for data analysis, to use R in a computer cluster. The proposed data analysis workflow is applied to a multi-source high dimensional drug discovery dataset and compared with a few existing R packages for parallel computing.

Arabidopsis SWI/SNF chromatin remodeling complex binds both promoters and terminators to regulate gene expression.

ATP-dependent chromatin remodeling complexes are important regulators of gene expression in Eukaryotes. In plants, SWI/SNF-type complexes have been shown critical for transcriptional control of key developmental processes, growth and stress responses. To gain insight into mechanisms underlying these roles, we performed whole genome mapping of the SWI/SNF catalytic subunit BRM in Arabidopsis thaliana, combined with transcript profiling experiments. Our data show that BRM occupies thousands of sites in Arabidopsis genome, most of which located within or close to genes. Among identified direct BRM transcriptional targets almost equal numbers were up- and downregulated upon BRM depletion, suggesting that BRM can act as both activator and repressor of gene expression. Interestingly, in addition to genes showing canonical pattern of BRM enrichment near transcription start site, many other genes showed a transcription termination site-centred BRM occupancy profile. We found that BRM-bound 3΄ gene regions have promoter-like features, including presence of TATA boxes and high H3K4me3 levels, and possess high antisense transcriptional activity which is subjected to both activation and repression by SWI/SNF complex. Our data suggest that binding to gene terminators and controlling transcription of non-coding RNAs is another way through which SWI/SNF complex regulates expression of its targets.

Atypical chronic myeloid leukaemia: A case of an orphan disease-A multicenter report by the Polish Adult Leukemia Group.

Atypical chronic myeloid leukaemia (aCML) belongs to myelodysplastic/myeloproliferative neoplasms. Because of its rarity and changing diagnostic criteria throughout subsequent classifications, data on aCML are very scarce. Therefore, we at the Polish Adult Leukemia Group performed a nationwide survey on aCML. Eleven biggest Polish centres participated in the study. Altogether, 45 patients were reported, among whom only 18 patients (40%) fulfilled diagnostic criteria. Among misdiagnosed patients, myelodysplastic/myeloproliferative syndrome unclassifiable and chronic myelomonocytic leukaemia were the most frequent diagnoses. Thirteen patients were male, median age 64.6 years (range 40.4-80.9). The median parameters at diagnosis were as follows: white blood cell count 97 × 109 /L (23.8-342) with immature progenitors amounting at 27.5% (12-72), haemoglobin 8.6 g/dL (3.9-14.9), and platelet count 66 × 109 /L (34-833). Cytoreductive treatment was used in all patients, and 2 patients underwent allogeneic hematopoietic stem cell transplantation. The median overall survival was 14.1 months (95% CI, 7.2), with median acute myeloid leukaemia-free survival of 13.3 months (95% CI, 3.6-22.6). Cumulative incidence of acute myeloid leukaemia transformation after 1 year in aCML group was 12.5% (95% CI, 0%-29.6%). To conclude, aCML harbours a poor prognosis. Treatment options are limited, with allogeneic hematopoietic stem cell transplantation being the only curative method at present, although only a minority of patients are transplant eligible. Educational measures are needed to improve the quality of diagnoses.

Nuclear pseudoinclusions in melanoma cells: prognostic fact or artifact? The possible role of Golgi phosphoprotein 3 overexpression in nuclear pseudoinclusions generation.

Nuclear pseudoinclusions (NPIs) are classically found in papillary thyroid carcinoma and meningioma. Although NPIs have been described in melanocytic lesions, there is no systematic analysis of potential relationship between NPIs and other clinicopathological characteristics of melanoma. We examined the presence of NPIs in H&E-stained tissue sections form 96 melanomas and analyzed statistical associations with important clinicopathological parameters and tissue immunoreactivity for selected proteins involved in epithelial-mesenchymal transition (SPARC, N-cadherin), cell adhesion and mobility (ALCAM, ADAM-10), regulation of mitosis (PLK1), cell survival (FOXP1) and functioning of Golgi apparatus (GOLPH3, GP73). NPIs were observed in 20% of melanomas and their presence correlated with high mitotic rate and ulceration of the tumor, but not with Breslow thickness, histologic type, or presence of metastases. We observed a significant correlation with shorter cancer-specific survival, but not disease-free survival. Presence of NPIs was related to high expression of GOLPH3 in melanoma cells, whereas their absence was linked to enhanced immunoreactivity of GOLPH3 in tumor-associated macrophages. NPIs are not an uncommon finding in skin melanoma and their diagnostic and prognostic utility could be helpful in the daily routine histopathological practice. The possible explanation of NPI generation is associated with enhanced activity of Golgi apparatus in melanoma cells.

Biological Aggressiveness of Subclinical No-Mass Ductal Carcinoma In Situ (DCIS) Can Be Reflected by the Expression Profiles of Epithelial-Mesenchymal Transition Triggers.

Epithelial-mesenchymal transitions (EMTs) have been recently implicated in the process of cancer progression. The aim of this study was to assess how the preoperative expression patterns of EMT biomarkers correlate with the risk of postoperative invasion in ductal carcinoma in situ (DCIS) found on stereotactic breast biopsies. N-cadherin, Snail1, and secreted protein acidic and rich in cysteine (SPARC) immunoreactivity was observed in 8%, 62%, and 38% of tumors, respectively. Snail1 and SPARC expressions were significantly related to N-cadherin expression and to each other. The postoperative upgrading rate was associated with a positive preoperative expression of all biomarkers. Significance of Snail1 and SPARC persisted in multivariate analysis, but the impact of SPARC on invasion was more significant. When these two EMT triggers were considered together, the risk of invasion did not significantly differ between the subtypes of DCIS with single positive expression (SPARC-/Snail1+ vs. SPARC+/Snail1-). However, it was significantly lower in single-positive DCIS when compared to lesions of a double-positive profile (SPARC+/Snail1+). Moreover, there were no cases in the double-negative DCIS (SPARC-/Snail1-), with foci of infiltrating cancer found postoperatively in residual postbiopsy lesions. In contrast, DCIS with a combined high SPARC and Snail1 expression (intermediate or strong) had an invasive component in 66⁻100% of tumors.

Epithelial-mesenchymal transition inducer Snail1 and invasive potential of intraductal breast cancer.

BACKGROUND AND OBJECTIVES: Transcription factor Snail1 is a key inducer of epithelial-mesenchymal transition (EMT), a biological process implicated in the cancer progression and metastasis. The aim of the study was to investigate Snail1 expression in DCIS found on breast biopsy and assess its predictive value for the final invasion. METHODS: A total of 209 patients with histologically diagnosed pure DCIS entered the study. Snail1 reactivity was evaluated with immunohistochemistry in tumor tissue from stereotactic vacuum-assisted biopsy of suspicious microcalcifications. RESULTS: Snail1 staining was observed in 62% of tumors: weak, intermediate, and strong in 27%, 21%, and 14% of lesions, respectively. Positive Snail1 expression was significantly rarer in DCIS presenting as powdery microcalcifications, when compared with crushed stone-like and casting-type and was more common in DCIS with comedonecrosis. Correlation with other features was not significant. None of standard parameters significantly influenced the upgrading rate. In contrast, in uni- and multivariate analysis the risk of postoperative invasion was significantly associated with positive Snail1 immunoreactivity. Moreover, there was a significant stepwise increase of upgrading rate according to Snail1 expression in DCIS cells: weak 9%, intermediate 26%, and strong 55%, respectively. CONCLUSIONS: Snail1 can reflect the invasive potential of DCIS and help identify its more aggressive subtypes.

Comparative analysis of LIN28-RNA binding sites identified at single nucleotide resolution.

It remains a formidable challenge to characterize the diverse complexes of RNA binding proteins and their targets. While crosslink and immunoprecipitation (CLIP) methods are powerful techniques that identify RNA targets on a global scale, the resolution and consistency of these methods is a matter of debate. Here we present a comparative analysis of LIN28-pre-let-7 UV-induced crosslinking using a tandem mass spectrometry (MS/MS) and deep sequencing interrogation of in vitro crosslinked complexes. Interestingly, analyses by the two methods diverge in their identification of crosslinked nucleotide identity - whereas bioinformatics and sequencing analyses suggest guanine in mammalian cells, MS/MS identifies uridine. This work suggests the need for comprehensive analysis and validation of crosslinking methodologies.

Ductal carcinoma in situ on stereotactic biopsy of suspicious breast microcalcifications: Expression of SPARC (Secreted Protein, Acidic and Rich in Cysteine) can predict postoperative invasion.

BACKGROUND AND OBJECTIVES: Secreted protein, acidic and rich in cysteine (SPARC) is able to play an important role in cancer invasion due to de-adhesive properties and impact on stromal remodeling. The aim of study was to investigate SPARC expression in preoperatively diagnosed breast DCIS and to assess its predictive value for the final invasion. METHODS: A total of 209 patients with DCIS found on stereotactic vacuum-assisted biopsy of suspicious microcalcifications were studied prospectively. RESULTS: SPARC staining was positive in luminal epithelial cells, stromal fibroblasts, and myoepithelial cells in 38%, 62%, and 61% of tumors, respectively. Neither patient age nor pattern of microcalcifications were related to SPARC expression. High nuclear grade and comedonecrosis were associated with strong immunoreactivity of SPARC in stromal fibroblasts and myoepithelial cells while not in luminal epithelial cells. Rate of postoperative invasion was significantly increased in DCIS with strong SPARC staining with regard to all investigated cells. None of standard parameters significantly influenced the upgrading risk. In multivariate analysis most significant and independent predictive factors were strong SPARC expression in luminal epithelial cells, and stromal fibroblasts. CONCLUSIONS: SPARC can be a new biomarker helpful to identify more aggressive DCIS and for prediction of invasive disease on final pathology. J. Surg. Oncol. 2016;114:548-556. © 2016 Wiley Periodicals, Inc.