A Deep Learning Decision Support Tool to Improve Risk Stratification and Reduce Unnecessary Biopsies in BI-RADS 4 Mammograms.

Purpose: To evaluate the performance of a biopsy decision support algorithmic model, the intelligent-augmented breast cancer risk calculator (iBRISK), on a multicenter patient dataset. Materials and Methods: iBRISK was previously developed by applying deep learning to clinical risk factors and mammographic descriptors from 9700 patient records at the primary institution and validated using another 1078 patients. All patients were seen from March 2006 to December 2016. In this multicenter study, iBRISK was further assessed on an independent, retrospective dataset (January 2015-June 2019) from three major health care institutions in Texas, with Breast Imaging Reporting and Data System (BI-RADS) category 4 lesions. Data were dichotomized and trichotomized to measure precision in risk stratification and probability of malignancy (POM) estimation. iBRISK score was also evaluated as a continuous predictor of malignancy, and cost savings analysis was performed. Results: The iBRISK model's accuracy was 89.5%, area under the receiver operating characteristic curve (AUC) was 0.93 (95% CI: 0.92, 0.95), sensitivity was 100%, and specificity was 81%. A total of 4209 women (median age, 56 years [IQR, 45-65 years]) were included in the multicenter dataset. Only two of 1228 patients (0.16%) in the "low" POM group had malignant lesions, while in the "high" POM group, the malignancy rate was 85.9%. iBRISK score as a continuous predictor of malignancy yielded an AUC of 0.97 (95% CI: 0.97, 0.98). Estimated potential cost savings were more than $420 million. Conclusion: iBRISK demonstrated high sensitivity in the malignancy prediction of BI-RADS 4 lesions. iBRISK may safely obviate biopsies in up to 50% of patients in low or moderate POM groups and reduce biopsy-associated costs.Keywords: Mammography, Breast, Oncology, Biopsy/Needle Aspiration, Radiomics, Precision Mammography, AI-augmented Biopsy Decision Support Tool, Breast Cancer Risk Calculator, BI-RADS 4 Mammography Risk Stratification, Overbiopsy Reduction, Probability of Malignancy (POM) Assessment, Biopsy-based Positive Predictive Value (PPV3) Supplemental material is available for this article. Published under a CC BY 4.0 license.See also the commentary by McDonald and Conant in this issue.

Clinical factors and association with treatment modalities in patients with breast cancer and brain metastases who develop leptomeningeal metastases.

PURPOSE: Leptomeningeal metastases (LM) are an aggressive complication of metastatic breast cancer (MBC) with brain metastases (BM), with a short survival of weeks to months. Studies suggest that surgical resection of BM may increase the risk of LM, especially in infratentorial metastases. In this retrospective study, we examine this and other factors which may be associated with increased risk of LM. METHODS: A database search at a single institution identified 178 patients with MBC and treated BM between 2007 and 2020. We collected demographic, clinical, radiographic, and other treatment data. LM was diagnosed by cerebrospinal fluid (CSF) cytology, neuroimaging, or both. Cox proportional hazards model was used. RESULTS: After a median follow-up of 8.5 months, 41 out of 178 patients (23%) with BM developed LM. Median time to develop LM was 130 days. Mean age was 51.3 years. The number and size of the BM, hemorrhagic/cystic lesions, progressive/stable systemic disease, and extracranial metastases sites other than liver did not pose a higher risk of LM. Infratentorial lesions (HR = 5.41) and liver metastases (HR = 2.28) had a higher risk of LM. Patients who had any surgery did not have a higher risk for LM (HR 1.13). The LM group had a worse overall survival as compared to the non-LM group. CONCLUSION: Among MBC patients with BM, infratentorial BM and visceral liver lesions increase the risk of LM, whereas local treatment modalities such as surgery and radiation do not. These data imply that local treatment strategy should not differ based on potential risk for LM.

Medical Complications and Prognostic Factors for Medically Unstable Conditions During Acute Inpatient Cancer Rehabilitation.

PURPOSE: Acute inpatient rehabilitation provides concurrent medical care and intensive rehabilitation. We sought to describe the nature of types of medical complications and to identify the more frequent types of medical complications management in patients with cancer undergoing acute inpatient rehabilitation. METHODS: We reviewed the records of all consecutive patients admitted to acute inpatient rehabilitation from September 1, 2017, through February 28, 2018. Presenting problem noted to be a significant change in medical status using the Centers for Medicare & Medicaid Services' Evaluation and Management Service Guide was defined as a medical complication. We recorded demographic and clinical characteristics. Multivariable logistic regression analysis was performed to assess prognostic factors for returning to acute care. RESULTS: Among 165 evaluable patients, 158 (96%) had at least one medical complication, and 31 (19%) had an unplanned return to acute care. After excluding three patients who had planned return to acute care, there was a cohort of 162 patients and the most common medical complication categories were electrolyte abnormalities 81 (50%), musculoskeletal 70 (43%), genitourinary or renal 61 (38%), hematologic 58 (36%), and cardiovascular problems 46 (28%). Multivariable analysis showed that tachycardia (odds ratio [OR], 7.83; 95% CI, 2.23 to 27.54; P = .001) and weekly or more frequent RBC transfusions (OR, 5.23; 95% CI, 1.39 to 19.64; P = .014) were independently associated with unplanned return to acute care. CONCLUSION: A high frequency and wide range of medical complications and interventions occur in patients with cancer undergoing acute inpatient rehabilitation. Close monitoring and expertise are needed for this patient population.

Intra-CSF topotecan in treatment of breast cancer patients with leptomeningeal metastases.

BACKGROUND: There are few treatment options for patients with leptomeningeal metastases (LM). METHODS: We report a case series of patients with breast cancer and LM treated with intra-CSF topotecan (TOPO). Outcome was assessed by clinical exam and MRI at baseline, at end of induction (4-5 weeks), then every 3 months; CSF cytology was determined at baseline and with each treatment. RESULTS: Thirty-one women [median age, 58 (37-81); median KPS 60 (40-100)] received treatment. At baseline, 68% had positive CSF cytology, and 90%, leptomeningeal enhancement on MRI. 84% of patients also received focal RT (not during TOPO) and 77% received concomitant systemic hormonal or chemotherapy. Median number of TOPO treatments was 14.5 (range, 3-71); median duration of treatment, 11 weeks (1-176); and median OS, 6.9 months (range, 0.9-48.8). Patients remaining progression-free during 4-6 weeks of induction (81%) had a median OS of 11.5 months (range, 1.8-48.8). Overall neurologic PFS at 6, 12, and 24 months was 39%, 26%, and 6%, respectively. Clearing of CSF malignant cells for >3 consecutive samples occurred in 10/21 (48%) patients with positive CSF cytology at baseline, remaining clear for a median duration of 15.9 months (range, 1.4-34.5). Grade 3 adverse events included headache or vomiting (3pts), T2 hyperintensity surrounding the ventricular catheter (2 pts), and meningitis (2 pts). CONCLUSIONS: Intra-CSF TOPO, with focal RT as needed for symptomatic areas of enhancement produced durable clearing of CSF malignant cells in 48% of patients positive at baseline, with promising median PFS and OS.

Metastatic triple-negative breast cancer: Established and emerging treatments.

Metastatic triple-negative breast cancer (mTNBC) patients tend to have a poor overall survival. The primary goals of treatment focus on palliation of symptoms and improvement in overall survival (OS). Single-agent sequential chemotherapy with anthracycline or taxane has remained the cornerstone of treatment for many years. The FDA has approved newer agents such as poly-adenosine diphosphate-ribose polymerase (PARP) inhibitors upfront in germline BRCA (gBRCA) 1/2 mutation carriers; atezolizumab and nab-paclitaxel combination frontline in patients with PD-L1 expression > 1%; and sacituzumab govitecan (IMMU-132), an antibody-drug conjugate in heavily pretreated mTNBC patients.

A randomized, controlled phase II trial of neoadjuvant ado-trastuzumab emtansine, lapatinib, and nab-paclitaxel versus trastuzumab, pertuzumab, and paclitaxel in HER2-positive breast cancer (TEAL study).

BACKGROUND: Neoadjuvant dual human epidermal growth factor receptor (HER2) blockade with trastuzumab and pertuzumab plus paclitaxel leads to an overall pathologic complete response (pCR) rate of 46%. Dual HER2 blockade with ado-trastuzumab emtansine (T-DM1) and lapatinib plus nab-paclitaxel has shown efficacy in patients with metastatic HER2-positive breast cancer. To test neoadjuvant effectiveness of this regimen, an open-label, multicenter, randomized, phase II trial was conducted comparing T-DM1, lapatinib, and nab-paclitaxel with trastuzumab, pertuzumab, and paclitaxel in patients with early-stage HER2-positive breast cancer. METHODS: Stratification by estrogen receptor (ER) status occurred prior to randomization. Patients in the experimental arm received 6 weeks of targeted therapies (T-DM1 and lapatinib) followed by T-DM1 every 3 weeks, lapatinib daily, and nab-paclitaxel weekly for 12 weeks. In the standard arm, patients received 6 weeks of trastuzumab and pertuzumab followed by trastuzumab weekly, pertuzumab every 3 weeks, and paclitaxel weekly for 12 weeks. The primary objective was to evaluate the proportion of patients with residual cancer burden (RCB) 0 or I. Key secondary objectives included pCR rate, safety, and change in tumor size at 6 weeks. Hypothesis-generating correlative assessments were also performed. RESULTS: The 30 evaluable patients were well-balanced in patient and tumor characteristics. The proportion of patients with RCB 0 or I was higher in the experimental arm (100% vs. 62.5% in the standard arm, p = 0.0035). In the ER-positive subset, all patients in the experimental arm achieved RCB 0-I versus 25% in the standard arm (p = 0.0035). Adverse events were similar between the two arms. CONCLUSION: In early-stage HER2-positive breast cancer, the neoadjuvant treatment with T-DM1, lapatinib, and nab-paclitaxel was more effective than the standard treatment, particularly in the ER-positive cohort. TRIAL REGISTRATION: Clinicaltrials.gov NCT02073487 , February 27, 2014.

Enhanced dermatologic toxicity following concurrent treatment with palbociclib and radiation therapy: A case report.

BACKGROUND: Cyclin-dependent kinase (CDK) 4/6 inhibitors represent a new class of targeted therapy options for the treatment of estrogen receptor-positive (ER+) human epidermal growth factor 2-negative (HER2-) metastatic breast cancer. There are currently no published prospective data on the safety of use of radiation treatment with palbociclib. CASE: We describe the case of a patient with metastatic breast cancer who received radiation treatment to a metastatic supraclavicular lymph node to planned 60 Gy in 30 fractions while on palbociclib, a selective inhibitor of CDK4/6. The patient developed early radiation toxicities including esophagitis and dermatitis that progressed to a severe left neck skin breakdown in the radiation field, resulting in the need for hospitalization. She had a break in treatment but was able to finish the radiation without palbociclib. Her tumor responded well to the treatment and her side effects healed. DISCUSSION: To our knowledge this is the first case to report on concurrent palbociclib and radiation use, with resultant enhanced radiation effects that required hospitalization for symptom management. Several preclinical studies have shown synergistic effects of radiation and both in vivo and in vitro experiments resulting in improved survival and decreased cell proliferation, respectively, through enhanced G1 cell cycle arrest. CONCLUSION: This case highlights the importance of using caution when combining radiation with the new targeted therapies. Until more data becomes available, physicians are recommended to exercise clinical judgment when deciding on whether to continue or discontinue a CDK4/6 inhibitor in a patient who may need radiation.

Correlating mammographic and pathologic findings in clinical decision support using natural language processing and data mining methods.

BACKGROUND: A key challenge to mining electronic health records for mammography research is the preponderance of unstructured narrative text, which strikingly limits usable output. The imaging characteristics of breast cancer subtypes have been described previously, but without standardization of parameters for data mining. METHODS: The authors searched the enterprise-wide data warehouse at the Houston Methodist Hospital, the Methodist Environment for Translational Enhancement and Outcomes Research (METEOR), for patients with Breast Imaging Reporting and Data System (BI-RADS) category 5 mammogram readings performed between January 2006 and May 2015 and an available pathology report. The authors developed natural language processing (NLP) software algorithms to automatically extract mammographic and pathologic findings from free text mammogram and pathology reports. The correlation between mammographic imaging features and breast cancer subtype was analyzed using one-way analysis of variance and the Fisher exact test. RESULTS: The NLP algorithm was able to obtain key characteristics for 543 patients who met the inclusion criteria. Patients with estrogen receptor-positive tumors were more likely to have spiculated margins (P = .0008), and those with tumors that overexpressed human epidermal growth factor receptor 2 (HER2) were more likely to have heterogeneous and pleomorphic calcifications (P = .0078 and P = .0002, respectively). CONCLUSIONS: Mammographic imaging characteristics, obtained from an automated text search and the extraction of mammogram reports using NLP techniques, correlated with pathologic breast cancer subtype. The results of the current study validate previously reported trends assessed by manual data collection. Furthermore, NLP provides an automated means with which to scale up data extraction and analysis for clinical decision support. Cancer 2017;114-121. © 2016 American Cancer Society.

Cell-free DNA as a molecular tool for monitoring disease progression and response to therapy in breast cancer patients.

Due to the spatial and temporal genomic heterogeneity of breast cancer, genomic sequencing obtained from a single biopsy may not capture the complete genomic profile of tumors. Thus, we propose that cell-free DNA (cfDNA) in plasma may be an alternate source of genomic information to provide comprehensive data throughout a patient's clinical course. We performed a retrospective chart review of 100 patients with stage 4 or high-risk stage 3 breast cancer. The degree of agreement between genomic alterations found in tumor DNA (tDNA) and cfDNA was determined by Cohen's Kappa. Clinical disease progression was compared to mutant allele frequency using a two-sided Fisher's exact test. The presence of mutations and mutant allele frequency was correlated with progression-free survival (PFS) using a Cox proportional hazards model and a log-rank test. The most commonly found genomic alterations were mutations in TP53 and PIK3CA, and amplification of EGFR and ERBB2. PIK3CA mutation and ERBB2 amplification demonstrated robust agreement between tDNA and cfDNA (Cohen's kappa = 0.64 and 0.77, respectively). TP53 mutation and EGFR amplification demonstrated poor agreement between tDNA and cfDNA (Cohen's kappa = 0.18 and 0.33, respectively). The directional changes of TP53 and PIK3CA mutant allele frequency were closely associated with response to therapy (p = 0.002). The presence of TP53 mutation (p = 0.0004) and PIK3CA mutant allele frequency [p = 0.01, HR 1.074 (95 % CI 1.018-1.134)] was excellent predictors of PFS. Identification of selected cancer-specific genomic alterations from cfDNA may be a noninvasive way to monitor disease progression, predict PFS, and offer targeted therapy.

Dual HER2 blockade: preclinical and clinical data.

The estrogen receptor and human epidermal growth factor receptor (HER) signaling pathways are the dominant drivers of cell proliferation and survival in the majority of human breast cancers. Not surprisingly, targeting these pathways provides the most effective therapies in appropriately selected patients. However, de novo and acquired resistance remain major obstacles to successful treatment. By increasing the understanding of the molecular mechanisms of combined HER2-targeted therapies, we aim to be better able to select patients who would respond to these treatments and understand some of the mechanisms of resistance to HER2-targeted treatments. Recent studies have demonstrated an increased effectiveness of dual targeted HER2 therapies against HER2-amplified breast cancer as compared with single blockade. These studies have resulted in the recent US Food and Drug Administration approval of the combination of taxane chemotherapy with pertuzumab and trastuzumab in the first-line metastatic setting as well as an accelerated approval in the neoadjuvant setting. Another mechanism for overcoming resistance to HER2 targeted therapies is the antibody-drug conjugate trastuzumab-emtansine, which targets the HER2 receptor conjugated to the potent antimicrotubule agent mertansine, allowing for intracellular release of the cytotoxic drug. Studies evaluating the efficacy of dual blockade with antibody-drug conjugate are currently ongoing. This article reviews recent data on different combinations of anti-HER2 treatments as well as ongoing and future research in this area.

Chemical castration of melanoma patients does not increase the frequency of tumor-specific CD4 and CD8 T cells after peptide vaccination.

Peptide vaccination against tumor-associated antigens remains one of the most common methods of immunization in cancer vaccine clinical trials. Although peptide vaccination has been reported to increase circulating antigen-specific T-cells, they have had limited clinical efficacy and there is a necessity to increase their capacity to generate strong antitumor responses. We sought to improve the clinical efficacy of peptide-based vaccines in cancer immunotherapy of metastatic melanoma using a LHRH agonist (leuprolide) as adjuvant. Seventy HLA-A*0201 stage IIb-IV melanoma patients were vaccinated with class I HLA-A*0201-restricted gp100209-2M peptide and stratified for HLA-DP4 restriction. HLA-DP4 patients were also vaccinated with class II HLA-DP4-restricted MAGE-3243-258 peptide. Patients from both groups were randomized to receive 2 doses of leuprolide or not. Here we report the increase in PBMC TREC levels at week 24 after peptide vaccination, which was independent of the leuprolide treatment. This change was mirrored by a small increase in the TREC-enriched CD8CD45RAROCD27CD103, but not the TREC-enriched CD4CD45RAROCD31 T-cell population. Serum concentration of 2 important factors for thymopoiesis was measured: insulin growth factor 1 (IGF-1) levels were not changed, whereas a moderate increase in IL-7 levels was noted in the sera of all patients 6 weeks after vaccination. Increased expression of CD127 (IL-7 receptor-α) at week 24, compared with baseline, was only seen in the CD8CD45RAROCD27CD103 T-cell population. Our results suggest that leuprolide has no effect on thymic output when used as peptide vaccine adjuvant, but IFA-based peptide vaccination may unexpectedly affect the thymus by increasing thymic output of new T cells.

Changes to adjuvant systemic therapy in breast cancer: a decade in review.

Breast cancer is the second most common cause of cancer death in women among the United States. Fortunately, it continues to be an active area of research. Today, it is well recognized that breast cancer can often be a systemic disease, with micrometastatic involvement at diagnosis in many patients. Over the past decade, adjuvant systemic therapy has been used to eradicate micrometastatic disease, and it has been shown to decrease the rates of recurrence and improve the survival of patients with early-stage, resected breast cancer. Some of the success of modern adjuvant systemic therapy has come from the advent of new chemotherapy and endocrine agents but also from the development of targeted therapies, which have improved the efficacy of conventional, cytotoxic therapy. There has also been increasing awareness that the dosing and schedule of administration of systemic therapies are equally important factors in achieving better outcomes in patients with early-stage breast cancer. Growing research into the biology and genomics of breast cancer has fueled the development of more accurate risk stratification tools and helped individualize the decision to recommend adjuvant systemic therapy. Herein, we present a review of salient developments over the past decade that have helped shape the adjuvant systemic therapy of today.

Breast cancer in Latinas: gene expression, differential response to treatments, and differential toxicities in Latinas compared with other population groups.

Disparities in clinical outcomes of breast cancer have been described among different racial and ethnic groups in the U.S. Convincing data exist showing that Latina women have a lower incidence of breast cancer but a higher breast cancer-related mortality rate compared with white women. Noticeable differences in breast cancer incidence are present even within different Latina subsets with a higher incidence in second- and third-generation women compared with foreign born. An increasing amount of data exists pointing to significant differences in the genetics and biology of breast cancer in Latinas as a significant contributor to the higher mortality, including a higher incidence of triple-negative breast cancers (which do not overexpress HER-2 protein and are negative for estrogen receptors and progesterone receptors). Other social and environmental factors are likely to play a significant role as well, including a lower rate of screening mammography, variable access to medical care, among others. Recent data are inconclusive regarding differences among racial/ethnic groups in the response to chemotherapy. Data on racial/ethnic variations in the pharmacogenomics of chemotherapy, endocrine treatments, and toxicity are more limited, with some data suggesting differences in frequencies of polymorphisms of genes involved in the metabolism of some of these agents. Further studies are needed on this subject.