Towards proactive palliative care in oncology: developing an explainable EHR-based machine learning model for mortality risk prediction.

BACKGROUND: Ex-ante identification of the last year in life facilitates a proactive palliative approach. Machine learning models trained on electronic health records (EHR) demonstrate promising performance in cancer prognostication. However, gaps in literature include incomplete reporting of model performance, inadequate alignment of model formulation with implementation use-case, and insufficient explainability hindering trust and adoption in clinical settings. Hence, we aim to develop an explainable machine learning EHR-based model that prompts palliative care processes by predicting for 365-day mortality risk among patients with advanced cancer within an outpatient setting. METHODS: Our cohort consisted of 5,926 adults diagnosed with Stage 3 or 4 solid organ cancer between July 1, 2017, and June 30, 2020 and receiving ambulatory cancer care within a tertiary center. The classification problem was modelled using Extreme Gradient Boosting (XGBoost) and aligned to our envisioned use-case: "Given a prediction point that corresponds to an outpatient cancer encounter, predict for mortality within 365-days from prediction point, using EHR data up to 365-days prior." The model was trained with 75% of the dataset (n = 39,416 outpatient encounters) and validated on a 25% hold-out dataset (n = 13,122 outpatient encounters). To explain model outputs, we used Shapley Additive Explanations (SHAP) values. Clinical characteristics, laboratory tests and treatment data were used to train the model. Performance was evaluated using area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC), while model calibration was assessed using the Brier score. RESULTS: In total, 17,149 of the 52,538 prediction points (32.6%) had a mortality event within the 365-day prediction window. The model demonstrated an AUROC of 0.861 (95% CI 0.856-0.867) and AUPRC of 0.771. The Brier score was 0.147, indicating slight overestimations of mortality risk. Explanatory diagrams utilizing SHAP values allowed visualization of feature impacts on predictions at both the global and individual levels. CONCLUSION: Our machine learning model demonstrated good discrimination and precision-recall in predicting 365-day mortality risk among individuals with advanced cancer. It has the potential to provide personalized mortality predictions and facilitate earlier integration of palliative care.

Outcomes in Nonmetastatic Hormone Receptor-Positive HER2-Negative Pure Mucinous Breast Cancer: A Multicenter Cohort Study.

BACKGROUND: Although considered a favorable subtype, pure mucinous breast cancer (PMBC) can recur, and evidence for adjuvant therapy is limited. We aimed to compare outcomes of nonmetastatic PMBC with invasive ductal carcinoma (IDC) and invasive lobular carcinoma (ILC) to address these uncertainties. METHODS: Individual patient-level data from 6 centers on stage I-III hormone receptor-positive and HER2-negative PMBC, IDC, and ILC were used to analyze recurrence-free interval (RFI), recurrence-free survival (RFS), and overall survival (OS), and to identify prognostic factors for PMBC. RESULTS: Data from 20,684 IDC cases, 1,475 ILC cases, and 943 PMBC cases were used. Median follow-up was 6.6 years. Five-year RFI, RFS, and OS for PMBC were 96.1%, 94.9%, and 98.1%, respectively. On multivariable Cox regression, PMBC demonstrated superior RFI (hazard ratio [HR], 0.59; 95% CI, 0.43-0.80), RFS (HR, 0.70; 95% CI, 0.56-0.89), and OS (HR, 0.71; 95% CI, 0.53-0.96) compared with IDC. ILC showed comparable outcomes to IDC. Fewer than half (48.7%) of recurrences in PMBC were distant, which was a lower rate than for IDC (67.3%) and ILC (80.6%). In contrast to RFI, RFS events were driven more by non-breast cancer deaths in older patients. Significant prognostic factors for RFI among PMBC included positive lymph node(s) (HR, 2.42; 95% CI, 1.08-5.40), radiotherapy (HR, 0.44; 95% CI, 0.23-0.85), and endocrine therapy (HR, 0.25; 95% CI, 0.09-0.70). No differential chemotherapy associations with outcomes were detected across PMBC subgroups by nodal stage, tumor size, and age. A separate SEER database analysis also did not find any association of improved survival with adjuvant chemotherapy in these subgroups. CONCLUSIONS: Compared with IDC, PMBC demonstrated superior RFI, RFS, and OS. Lymph node negativity, adjuvant radiotherapy, and endocrine therapy were associated with superior RFI. Adjuvant chemotherapy was not associated with better outcomes.

Determining the Need for Metastatic Staging in Patients with Bilateral Breast Cancers.

Introduction: Bilateral breast cancers (BBC) diagnosed at an interval apart are uncommon. While metastatic staging guidelines are established in patients with unilateral breast cancer, its role in BBC diagnosed at an interval apart is unclear. We aim to identify the subgroup who would benefit from metastatic staging at contralateral cancer diagnosis. Methods: Eligible patients were divided into three categories: (A) ipsilateral invasive cancer and contralateral ductal carcinoma in situ (DCIS), (B) bilateral invasive cancers and (C) ipsilateral DCIS and contralateral invasive cancer and reviewed retrospectively. We excluded patients with bilateral DCIS, synchronous BBC diagnosed within 6 months from first cancer, patients who were stage IV at first cancer diagnosis and patients with recurrence prior to contralateral cancer. Results: Of 4516 newly diagnosed breast cancer patients, 79 patients were included. Systemic metastasis occurred in 15.6% of patients in Group B. Having nodal positivity of either cancer which were diagnosed ≤30 months apart and nodal positivity of only the contralateral cancer when diagnosed >30 months apart was significantly associated with systemic metastasis (p = 0.0322). Conclusions: Both the nodal status and a 30 months cut-off time interval between the two cancers can be used to identify patients who will benefit from metastatic staging. This finding requires validation in larger studies.

The cardioprotective potential of sodium-glucose cotransporter 2-inhibitors in breast cancer therapy-related cardiac dysfunction - A systematic review.

BACKGROUND: Sodium-glucose cotransporter 2-inhibitors (SGLT2i) improve cardiovascular outcomes including reduction in risk of first hospitalisation for heart failure (HF), worsening HF and cardiovascular death regardless of HF or diabetes mellitus (DM) status. It is not known whether SGLT2i can prevent the development of incident HF or reduce the risk of HF in patients receiving trastuzumab with or without other concurrent anti-HER2 agent or sequential anthracycline for treatment of HER2 positive breast cancer. Patients with active malignancy or recent history of malignancy were excluded from participating in the main cardiovascular outcome trials involving SGLT2i. AIM: A systematic review was performed to objectively assess published literature on the cardioprotective effects of SGLT2i in breast cancer treatment-related cardiotoxicity. METHODS: Systematic searches of Embase, Medline, The Cochrane Central Register of Controlled Trials (CENTRAL), and ClinicalTrials.gov databases were performed. Titles and abstracts were screened separately by two cardio-oncologists (JHC, WTC). Full texts of potentially eligible records were then assessed separately by JHC and WTC before inclusion into review upon joint agreement. RESULTS: 479 records were identified from 3 databases (MEDLINE=51, EMBASE=408, CENTRAL=13) and 1 registry (Clinicaltrials.gov=7). 460 records were excluded based on title and abstract (including duplicates). 19 full text reports were assessed for eligibility and included in review (basic science/animal study paper 2, Clinicaltrials.gov randomised controlled trial submission 1 (currently recruiting), basic science/animal study conference abstract 5, case report 2, review 3, editorial comment 2, clinical guidelines 1, retrospective/registry-based conference abstract 3). CONCLUSION: Cardiotoxicity is the most common dose-limiting toxicity associated with trastuzumab. Discontinuation of trastuzumab however, can lead to worse cancer outcomes. There have been case reports, registry-based, retrospective cohort-based and mechanistic studies suggesting the cardioprotective potential of SGLT2i in cancer therapy-related cardiac dysfunction (CTRCD). Based on these, there is now a call for randomised controlled trials to be performed in this patient cohort to advise guideline-directed therapy for CTRCD, which will in turn also provide detailed safety information and improve cancer and cardiovascular outcomes.

How Asian Breast Cancer Patients Experience Unequal Incidence of Chemotherapy Side Effects: A Look at Ethnic Disparities in Febrile Neutropenia Rates.

The majority of published findings on chemotherapy-induced febrile neutropenia (FN) are restricted to three ethnic groups: Asians, Caucasians, and African Americans. In this two-part study, we examined FN incidence and risk factors in Chinese, Malay, and Indian chemotherapy-treated breast cancer (BC) patients. Hospital records or ICD codes were used to identify patients with FN. In both the Singapore Breast Cancer Cohort (SGBCC) and the Joint Breast Cancer Registry (JBCR), the time of the first FN from the start of chemotherapy was estimated using Cox regression. Multinomial regression was used to evaluate differences in various characteristics across ethnicities. FN was observed in 170 of 1014 patients in SGBCC. The Cox model showed that non-Chinese were at higher risk of developing FN (HRMalay [95% CI]:2.04 [1.44-2.88], p < 0.001; HRIndian:1.88 [1.11-3.18], p = 0.018). In JBCR, FN was observed in 965 of 7449 patients. Univariable Cox models identified ethnicity, a lower baseline absolute neutrophil count, non-luminal A proxy subtypes, and anthracycline-containing regimens as risk factors. Disparities across ethnicities' risk (HRMalay:1.29 [1.07-1.54], p = 0.006; HRIndian:1.50 [1.19-1.88], p < 0.001) remained significant even after further adjustments. Finally, an age-adjusted multinomial model showed that Malays (p = 0.006) and Indians (p = 0.009) were significantly more likely to develop multiple episodes of FN during treatment. Ethnic differences in chemotherapy-induced FN among BC patients exist. Further studies can focus on investigating pharmacogenetic differences across ethnicities.

Inferring cancer disease response from radiology reports using large language models with data augmentation and prompting.

OBJECTIVE: To assess large language models on their ability to accurately infer cancer disease response from free-text radiology reports. MATERIALS AND METHODS: We assembled 10 602 computed tomography reports from cancer patients seen at a single institution. All reports were classified into: no evidence of disease, partial response, stable disease, or progressive disease. We applied transformer models, a bidirectional long short-term memory model, a convolutional neural network model, and conventional machine learning methods to this task. Data augmentation using sentence permutation with consistency loss as well as prompt-based fine-tuning were used on the best-performing models. Models were validated on a hold-out test set and an external validation set based on Response Evaluation Criteria in Solid Tumors (RECIST) classifications. RESULTS: The best-performing model was the GatorTron transformer which achieved an accuracy of 0.8916 on the test set and 0.8919 on the RECIST validation set. Data augmentation further improved the accuracy to 0.8976. Prompt-based fine-tuning did not further improve accuracy but was able to reduce the number of training reports to 500 while still achieving good performance. DISCUSSION: These models could be used by researchers to derive progression-free survival in large datasets. It may also serve as a decision support tool by providing clinicians an automated second opinion of disease response. CONCLUSIONS: Large clinical language models demonstrate potential to infer cancer disease response from radiology reports at scale. Data augmentation techniques are useful to further improve performance. Prompt-based fine-tuning can significantly reduce the size of the training dataset.

Breast cancer risk stratification using genetic and non-genetic risk assessment tools for 246,142 women in the UK Biobank.

PURPOSE: The benefit of using individual risk prediction tools to identify high-risk individuals for breast cancer (BC) screening is uncertain, despite the personalized approach of risk-based screening. METHODS: We studied the overlap of predicted high-risk individuals among 246,142 women enrolled in the UK Biobank. Risk predictors assessed include the Gail model (Gail), BC family history (FH, binary), BC polygenic risk score (PRS), and presence of loss-of-function (LoF) variants in BC predisposition genes. Youden J-index was used to select optimal thresholds for defining high-risk. RESULTS: In total, 147,399 were considered at high risk for developing BC within the next 2 years by at least 1 of the 4 risk prediction tools examined (Gail2-year > 0.5%: 47%, PRS2-yea r > 0.7%: 30%, FH: 6%, and LoF: 1%); 92,851 (38%) were flagged by only 1 risk predictor. The overlap between individuals flagged as high-risk because of genetic (PRS) and Gail model risk factors was 30%. The best-performing combinatorial model comprises a union of high-risk women identified by PRS, FH, and, LoF (AUC2-year [95% CI]: 62.2 [60.8 to 63.6]). Assigning individual weights to each risk prediction tool increased discriminatory ability. CONCLUSION: Risk-based BC screening may require a multipronged approach that includes PRS, predisposition genes, FH, and other recognized risk factors.

Will Absolute Risk Estimation for Time to Next Screen Work for an Asian Mammography Screening Population?

Personalized breast cancer risk profiling has the potential to promote shared decision-making and improve compliance with routine screening. We assessed the Gail model's performance in predicting the short-term (2- and 5-year) and the long-term (10- and 15-year) absolute risks in 28,234 asymptomatic Asian women. Absolute risks were calculated using different relative risk estimates and Breast cancer incidence and mortality rates (White, Asian-American, or the Singapore Asian population). Using linear models, we tested the association of absolute risk and age at breast cancer occurrence. Model discrimination was moderate (AUC range: 0.580-0.628). Calibration was better for longer-term prediction horizons (E/Olong-term ranges: 0.86-1.71; E/Oshort-term ranges:1.24-3.36). Subgroup analyses show that the model underestimates risk in women with breast cancer family history, positive recall status, and prior breast biopsy, and overestimates risk in underweight women. The Gail model absolute risk does not predict the age of breast cancer occurrence. Breast cancer risk prediction tools performed better with population-specific parameters. Two-year absolute risk estimation is attractive for breast cancer screening programs, but the models tested are not suitable for identifying Asian women at increased risk within this short interval.

Pregnancy-associated breast cancer: a multicenter study comparing clinicopathological factors, diagnosis and treatment outcomes with non-pregnant patients.

PURPOSE: Pregnancy-associated breast cancer (PABC), defined as breast carcinoma diagnosed during pregnancy or in the first post-partum year, is one of the most common gestation-related malignancies with reported differences in tumor characteristics and outcomes. This multicenter study aims to review cases of PABC in Singapore, including their clinicopathological features, treatment, and clinical outcomes compared to non-PABC patients. METHODS: Demographic, histopathologic and clinical outcomes of 93 PABC patients obtained from our database were compared to 1424 non-PABC patients. RESULTS: PABC patients presented at a younger age. They had higher tumor and nodal stages, higher tumor grade, were more likely to be hormone receptor negative and had a higher incidence of multicentric and multifocal tumors. Histological examination after definitive surgery showed no significant difference in tumor size and number of positive lymph nodes suggesting similar neoadjuvant treatment effects. Despite this, PABC patients had worse outcomes with poorer overall survival and disease-free survival, OS (P < 0.0001) and DFS (P < 0.0001). Termination of pregnancy did not improve survival. CONCLUSION: Patients with PABC present at a higher stage with more aggressive disease and have poorer outcomes compared to non-PABC patients. Reducing delay in diagnosis and treatment may help improve survival.

Is Metastatic Staging Needed for All Patients with Synchronous Bilateral Breast Cancers?

BACKGROUND: Patients with bilateral breast cancers are uncommon and are associated with a poorer prognosis. While metastatic staging guidelines in patients with unilateral cancer were established, the indication of metastatic staging in patients with bilateral breast cancers is unclear. We aimed to determine which patients with synchronous bilateral breast cancers require metastatic staging at diagnosis. This is the first such reported study, to the best of our knowledge. METHODS: A retrospective review of newly diagnosed synchronous bilateral invasive breast cancer patients at our institution was performed. We excluded patients with malignant phyllodes or no metastatic staging. Patients' demographics and pathological and staging results were analysed to determine the group of bilateral breast cancer patients who required metastatic staging. RESULTS: A total of 92 patients with synchronous bilateral invasive cancers were included. The mean age was 58 years old, and 64.1% had bilateral invasive ductal carcinoma. 23.9% had systemic metastasis. Nodal status was statistically significant for systemic metastasis on staging (p = 0.0081), with only three patients (3.3%) having negative nodal status and positive metastatic staging. These three patients, however, showed symptoms of distant metastasis. 92.3% of patients with negative nodes also had negative metastatic staging. Using negative nodal status as a guide avoided metastatic staging in 40.4% of all patients. CONCLUSIONS: Negative nodal status was the most predictive factor for no systemic metastasis on staging in patients with synchronous bilateral invasive breast cancers. Hence, metastatic staging could be reserved for patients with symptoms of systemic metastasis and/or metastatic nodes. This finding could be validated in larger studies.

Association between Breast Cancer Polygenic Risk Score and Chemotherapy-Induced Febrile Neutropenia: Null Results.

BACKGROUND: The hypothesis that breast cancer (BC) susceptibility variants are linked to chemotherapy-induced toxicity has been previously explored. Here, we investigated the association between a validated 313-marker-based BC polygenic risk score (PRS) and chemotherapy-induced neutropenia without fever and febrile neutropenia (FNc) in Asian BC patients. METHODS: This observational case-control study of Asian BC patients treated with chemotherapy included 161 FNc patients, 219 neutropenia patients, and 936 patients who did not develop neutropenia. A continuous PRS was calculated by summing weighted risk alleles associated with overall, estrogen receptor- (ER-) positive, and ER-negative BC risk. PRS distributions neutropenia or FNc cases were compared to controls who did not develop neutropenia using two-sample t-tests. Odds ratios (OR) and corresponding 95% confidence intervals were estimated for the associations between PRS (quartiles and per standard deviation (SD) increase) and neutropenia-related outcomes compared to controls. RESULTS: PRS distributions were not significantly different in any of the comparisons. Higher PRSoverall quartiles were negatively correlated with neutropenia or FNc. However, the associations were not statistically significant (PRS per SD increase OR neutropenia: 0.91 [0.79-1.06]; FNc: 0.87 [0.73-1.03]). No dose-dependent trend was observed for the ER-positive weighted PRS (PRSER-pos) and ER-negative weighted PRS (PRSER-neg). CONCLUSION: BC PRS was not strongly associated with chemotherapy-induced neutropenia or FNc.

Modern-era radiotherapy and ischaemic heart disease-related mortality outcomes in Asian breast-cancer patients.

Introduction: Left-sided breast-cancer patients treated with adjuvant radiotherapy (RT) before the 1990s were associated with increased risk of cardiac mortality. Modern RT techniques have since improved, resulting in lower radiation doses to the heart. However, concerns regarding cardiac toxicity remain. In a retrospective cohort study, we compare the ischaemic heart disease (IHD)-related mortality of left-sided versus right-sided breast-cancer patients. We present the results of the cardiac mortality and all-cause mortality risk of Asian breast-cancer survivors treated with RT in Singapore. Material and methods: A total of 14,419 Asian women from a single institution were treated for breast cancer from 2000 to 2016. A systematic mortality follow-up was conducted until December 2015. The effect of breast cancer laterality on IHD-related mortality and on overall mortality was investigated. Mean heart doses were recorded for patients from 2010-2016. Results: In the irradiated group (n = 9556), we found no difference in IHD-related mortality or overall mortality when comparing the left- and right-sided breast cancers. The hazard ratio of cardiac mortality for left-sided versus right-sided RT was 0.94 (95% CI: 0.64-1.38). The hazard ratio for all-cause mortality was 1.03 (95% CI: 0.94-1.13). Conclusions: Our study of Asian cancer patients did not reveal a significant increase in the risk of IHD-related mortality or overall mortality comparing left- vs. right-sided breast cancers in modern-era RT.

The Impact of Statin Use and Breast Cancer Recurrence - A Retrospective Study in Singapore.

Introduction: Statins, HMG-CoA reductase inhibitors, are commonly used cholesterol-lowering medications which are also increasingly recognized to have anti-cancer properties for various cancers, including breast cancer. Most clinical evidence supports a protective effect of statin on reducing breast cancer recurrence, particularly in hormone-receptor positive breast cancers.This study seeks to study the impact of statin use on breast cancer recurrence in an Asian population. Methods: This is a retrospective study of patients diagnosed with breast cancer at the National Cancer Centre and Singapore General Hospital from 2005-2015. Statin use was defined as use after surgery. Associations between statin use, breast cancer recurrence and overall survival were estimated using Cox proportional hazards regression with adjustment for age, TNM stage, grade, ER/HER2 status, and co-morbidities. Associations between statin-use and disease-specific survival were estimated using competing risks regression. Results: A total of 7858 females with breast cancer were studied, 1353(17.2%) were statin users, 6505(82.8%) were non-statin users, with a median follow-up of 8.67 years. Distribution of cancer stage, histology, molecular subtypes and grades were similar in both groups. Estrogen receptor(ER) positive (HR 0.57,95%CI 0.43-0.76,p<0.001) and HER2 negative (HR 0.74,95%CI 0.57-0.96,p=0.026) invasive cancers had a lower risk of recurrence in statin users. Statin users trended towards a long term recurrence-risk reduction (all subtypes,HR 0.48,p=0.002; ER-, HR 0.34,p=0.036; HER2+,HR 0.10,p=0.002). The risk-reduction benefit is not appreciated in statin users with DCIS, possibly due to small recurrence event numbers. Disease-specific survival benefit was seen in statin users with ER+ cancers (adjusted SHR 0.71,95%CI 0.53-0.96,p=0.027), especially ER+ invasive cancers (adjusted SHR 0.72, 95%CI 0.53-0.97,p=0.028), but with no statistically significant benefit in overall survival for statin users (all subtypes). Conclusion: This is the first known retrospective study on the effect of statin use and breast cancer recurrence in an Asian population. Similar to previous international studies, statin use is associated with a risk reduction in breast cancer recurrence. This is especially beneficial in patients who have ER+ and HER2- invasive breast cancer. Statin use is also associated with a reduced risk of breast cancer recurrence in all subtypes of breast cancer in the long term (>6 years post diagnosis).

Male breast cancer: a Singapore perspective.

INTRODUCTION: Male breast cancer (MBC) is rare, representing <1% of all breast cancers. Treatment recommendations have been extrapolated from trial data of female breast cancer patients. This study aims to report our institutional experience of MBC across a 20 year period, analyse the survival outcome and prognosis of this group against female breast cancer patients treated at the same centre. METHODS: Clinical, histopathological, treatment and survival data of male and female breast cancer patients treated between Jan 1999 and July 2019 at Singapore General Hospital and National Cancer Centre Singapore were identified and analysed. RESULTS: Fifty-seven male patients were identified. The median age at diagnosis was 63 years. Majority had invasive ductal carcinoma (86%) and presented at an early disease stage: 70.2% presented as Tis/T1/T2 and 49.1% had no axillary nodal involvement. 84.2% had a simple mastectomy with either a sentinel lymph node biopsy or axillary clearance. The median follow up was 5.69 years for males and 5.83 years for females. The median survival was 11.86 years for males and 16.3 years for females. At 5 years, overall survival (OS) was 69.9% (52.3-82.1%) and disease free survival (DFS) was 62.9% (44.9-76.5%) for males compared with OS 83.8% (83.21-84.39%) and DFS 74.5% (73.91-75.09%) for females. CONCLUSION: MBC remains understudied. Our institutional data indicates that good long term survival in South-East Asian patients can be achieved with treatment protocols that are similar to female breast cancer. More prospective studies are required.

Overlap of high-risk individuals predicted by family history, and genetic and non-genetic breast cancer risk prediction models: implications for risk stratification.

BACKGROUND: Family history, and genetic and non-genetic risk factors can stratify women according to their individual risk of developing breast cancer. The extent of overlap between these risk predictors is not clear. METHODS: In this case-only analysis involving 7600 Asian breast cancer patients diagnosed between age 30 and 75 years, we examined identification of high-risk patients based on positive family history, the Gail model 5-year absolute risk [5yAR] above 1.3%, breast cancer predisposition genes (protein-truncating variants [PTV] in ATM, BRCA1, BRCA2, CHEK2, PALB2, BARD1, RAD51C, RAD51D, or TP53), and polygenic risk score (PRS) 5yAR above 1.3%. RESULTS: Correlation between 5yAR (at age of diagnosis) predicted by PRS and the Gail model was low (r=0.27). Fifty-three percent of breast cancer patients (n=4041) were considered high risk by one or more classification criteria. Positive family history, PTV carriership, PRS, or the Gail model identified 1247 (16%), 385 (5%), 2774 (36%), and 1592 (21%) patients who were considered at high risk, respectively. In a subset of 3227 women aged below 50 years, the four models studied identified 470 (15%), 213 (7%), 769 (24%), and 325 (10%) unique patients who were considered at high risk, respectively. For younger women, PRS and PTVs together identified 745 (59% of 1276) high-risk individuals who were not identified by the Gail model or family history. CONCLUSIONS: Family history and genetic and non-genetic risk stratification tools have the potential to complement one another to identify women at high risk.

HER2 expression, copy number variation and survival outcomes in HER2-low non-metastatic breast cancer: an international multicentre cohort study and TCGA-METABRIC analysis.

BACKGROUND: HER2-low breast cancer (BC) is currently an area of active interest. This study evaluated the impact of low expression of HER2 on survival outcomes in HER2-negative non-metastatic breast cancer (BC). METHODS: Patients with HER2-negative non-metastatic BC from 6 centres within the Asian Breast Cancer Cooperative Group (ABCCG) (n = 28,280) were analysed. HER2-low was defined as immunohistochemistry (IHC) 1+ or 2+ and in situ hybridization non-amplified (ISH-) and HER2-zero as IHC 0. Relapse-free survival (RFS) and overall survival (OS) by hormone receptor status and HER2 IHC 0, 1+ and 2+ ISH- status were the main outcomes. A combined TCGA-BRCA and METABRIC cohort (n = 1967) was also analysed to explore the association between HER2 expression, ERBB2 copy number variation (CNV) status and RFS. RESULTS: ABCCG cohort median follow-up was 6.6 years; there were 12,260 (43.4%) HER2-low BC and 16,020 (56.6%) HER2-zero BC. The outcomes were better in HER2-low BC than in HER2-zero BC (RFS: centre-adjusted hazard ratio (HR) 0.88, 95% CI 0.82-0.93, P < 0.001; OS: centre-adjusted HR 0.82, 95% CI 0.76-0.89, P < 0.001). On multivariable analysis, HER2-low status was prognostic (RFS: HR 0.90, 95% CI 0.85-0.96, P = 0.002; OS: HR 0.86, 95% CI 0.79-0.93, P < 0.001). These differences remained significant in hormone receptor-positive tumours and for OS in hormone receptor-negative tumours. Superior outcomes were observed for HER2 IHC1+ BC versus HER2-zero BC (RFS: HR 0.89, 95% CI 0.83-0.96, P = 0.001; OS: HR 0.85, 95% CI 0.78-0.93, P = 0.001). No significant differences were seen between HER2 IHC2+ ISH- and HER2-zero BCs. In the TCGA-BRCA and METABRIC cohorts, ERBB2 CNV status was an independent RFS prognostic factor (neutral versus non-neutral HR 0.71, 95% CI 0.59-0.86, P < 0.001); no differences in RFS by ERBB2 mRNA expression levels were found. CONCLUSIONS: HER2-low BC had a superior prognosis compared to HER2-zero BC in the non-metastatic setting, though absolute differences were modest and driven by HER2 IHC 1+ BC. ERBB2 CNV merits further investigation in HER2-negative BC.

Germline breast cancer susceptibility genes, tumor characteristics, and survival.

BACKGROUND: Mutations in certain genes are known to increase breast cancer risk. We study the relevance of rare protein-truncating variants (PTVs) that may result in loss-of-function in breast cancer susceptibility genes on tumor characteristics and survival in 8852 breast cancer patients of Asian descent. METHODS: Gene panel sequencing was performed for 34 known or suspected breast cancer predisposition genes, of which nine genes (ATM, BRCA1, BRCA2, CHEK2, PALB2, BARD1, RAD51C, RAD51D, and TP53) were associated with breast cancer risk. Associations between PTV carriership in one or more genes and tumor characteristics were examined using multinomial logistic regression. Ten-year overall survival was estimated using Cox regression models in 6477 breast cancer patients after excluding older patients (≥75years) and stage 0 and IV disease. RESULTS: PTV9genes carriership (n = 690) was significantly associated (p < 0.001) with more aggressive tumor characteristics including high grade (poorly vs well-differentiated, odds ratio [95% confidence interval] 3.48 [2.35-5.17], moderately vs well-differentiated 2.33 [1.56-3.49]), as well as luminal B [HER-] and triple-negative subtypes (vs luminal A 2.15 [1.58-2.92] and 2.85 [2.17-3.73], respectively), adjusted for age at diagnosis, study, and ethnicity. Associations with grade and luminal B [HER2-] subtype remained significant after excluding BRCA1/2 carriers. PTV25genes carriership (n = 289, excluding carriers of the nine genes associated with breast cancer) was not associated with tumor characteristics. However, PTV25genes carriership, but not PTV9genes carriership, was suggested to be associated with worse 10-year overall survival (hazard ratio [CI] 1.63 [1.16-2.28]). CONCLUSIONS: PTV9genes carriership is associated with more aggressive tumors. Variants in other genes might be associated with the survival of breast cancer patients. The finding that PTV carriership is not just associated with higher breast cancer risk, but also more severe and fatal forms of the disease, suggests that genetic testing has the potential to provide additional health information and help healthy individuals make screening decisions.

Breast cancer risk stratification for mammographic screening: A nation-wide screening cohort of 24,431 women in Singapore.

BACKGROUND: Breast cancer incidence is increasing in Asia. However, few women in Singapore attend routine mammography screening. We aim to identify women at high risk of breast cancer who will benefit most from regular screening using the Gail model and information from their first screen (recall status and mammographic density). METHODS: In 24,431 Asian women (50-69 years) who attended screening between 1994 and 1997, 117 developed breast cancer within 5 years of screening. Cox proportional hazard models were used to study the associations between risk classifiers (Gail model 5-year absolute risk, recall status, mammographic density), and breast cancer occurrence. The efficacy of risk stratification was evaluated by considering sensitivity, specificity, and the proportion of cancers identified. RESULTS: Adjusting for information from first screen attenuated the hazard ratios (HR) associated with 5-year absolute risk (continuous, unadjusted HR [95% confidence interval]: 2.3 [1.8-3.1], adjusted HR: 1.9 [1.4-2.6]), but improved the discriminatory ability of the model (unadjusted AUC: 0.615 [0.559-0.670], adjusted AUC: 0.703 [0.653-0.753]). The sensitivity and specificity of the adjusted model were 0.709 and 0.622, respectively. Thirty-eight percent of all breast cancers were detected in 12% of the study population considered high risk (top five percentile of the Gail model 5-year absolute risk [absolute risk ≥1.43%], were recalled, and/or mammographic density ≥50%). CONCLUSION: The Gail model is able to stratify women based on their individual breast cancer risk in this population. Including information from the first screen can improve prediction in the 5 years after screening. Risk stratification has the potential to pick up more cancers.

Outcomes of Stage I and II Breast Cancer with Nodal Micrometastases Treated with Mastectomy without Axillary Therapy.

PURPOSE: Studies that report equivalent oncologic outcomes of sentinel lymph node biopsy (SLNB) alone versus axillary lymph node dissection (ALND) for T1-2N1mi breast cancers are heavily weighted with patients who received breast-conserving surgery (BCS). The impact of omitting ALND in N1mi patients treated with mastectomy is not well studied. It is also unknown if these patients would benefit from post-mastectomy radiotherapy (PMRT). This study reports the outcomes of patients with T1-2N1mi breast cancer treated by mastectomy without axillary therapy. METHODS: Patients who had T1-2N1mi breast cancer and underwent mastectomy from January 1998 to December 2018 were identified from our multi-institutional prospective database. Axillary recurrence rate (ARR), disease-free survival (DFS), and overall survival (OS) are reported. RESULTS: 260 patients with pT1-2N1mi breast cancer who had mastectomy were identified. They had either SLNB (35.4%) or ALND (64.6%). Majority of these patients received adjuvant systemic therapy (93.8%). 77 (29.6%) patients received radiotherapy, 31 after SLNB and 46 after ALND. At median follow-up of 61 months, ARR was 1.1% (n = 1) in the SLNB only group, vs. 0.6% (n = 1) in the ALND group (p = 0.752). DFS and OS were not significantly different between patients with SLNB alone versus ALND (p = 0.40 and p = 0.27, respectively). Among 92 patients who had SLNB only, no DFS or OS difference was observed with the use of PMRT. CONCLUSION: In T1-2N1mi patients with mastectomy and SLNB, axillary recurrences were rare. No statistically significant differences were noted between patients with SLNB, ALND, or PMRT. Our findings suggest that these patients may be safely treated without axillary therapy.

A Prospective Randomized Controlled Trial to Compare the Use of Conventional Dark-Ink Tattoo and Ultraviolet-Ink Tattoo for Patients Undergoing Breast Radiation Therapy.

PURPOSE: Permanent tattoo marks used in radiation therapy remain for the duration of treatment and essentially for the rest of the patient's life. This study compared the initial positioning setup errors and body image perception between patients with ultraviolet (UV) and conventional dark ink tattoos. METHODS AND MATERIALS: Thirty-four patients from February 2018 to March 2019, who underwent radiation therapy (RT) to the breast or chest wall for ductal carcinoma in situ or breast cancer were prospectively recruited and randomized (1:1) to receive either conventional dark ink or UV ink tattoos. Each patient received the assigned tattoos during computed tomography (CT) simulation and initial treatment setup shifts were compared. A 9-item body-image survey was administered to all patients at 3 time points: CT simulation, last week of RT, and 6 weeks post-RT. Feedback from CT and treatment staff in terms of setup time and challenges were collated. RESULTS: The median age of the patient cohort was 46 years old. No statistically significant difference was observed between the mean setup errors for the conventional dark ink group (0.11 cm inferior, 0.01 cm left, 0.11 cm posterior) and UV ink group (0.01 cm superior, 0.01 cm right, 0.06 cm posterior; P = NS). Similar responses were observed in the body-image survey between the 2 groups across all time points (P = NS). The majority of the patients (dark ink 82.3% vs UV ink 88.2%) did not feel less sexually attractive as a result of the tattoo at 6 weeks post-RT. At 6 weeks post-RT, patients in both groups were satisfied with the appearance of the tattoo and did not feel cautious about their choice of clothes (82.4% vs 88.2%; P = NS). In addition, 88.6% of staff (n = 35) felt minimum effect of UV ink on the overall setup time, and 94.3% found no difficulty localizing the UV ink tattoos during patient positioning. CONCLUSIONS: No difference in setup accuracy was found using UV ink tattoos, and it could be implemented clinically with minimal effect on the existing workflow. Patients expressed high satisfaction and self-confidence with the use of UV ink tattoos.