Dave P. Miller

Out-of-hospital cardiac arrest is a time-sensitive emergency that requires prompt identification and intervention: sudden, unwitnessed cardiac arrest is nearly unsurvivable^1–3. A cardinal sign of cardiac arrest is sudden loss of pulse⁴. Automated biosensor detection of unwitnessed cardiac arrest, and dispatch of medical assistance, may improve survivability given the significant prognostic role of time^3,5, but only if the false positive burden on public emergency medical systems is minimized^5–7. Here we show that a multimodal, machine learning-based algorithm on a smartwatch can reach performance thresholds making it deployable at societal scale. First, using photoplethysmography (PPG), we show that wearable PPG measurements of peripheral pulselessness (induced via an arterial occlusion model) manifest similarly to pulselessness caused by a common cardiac arrest arrhythmia, ventricular fibrillation (VF). Based on the similarity of the PPG signal (from VF or arterial occlusion), we developed and validated a loss of pulse detection algorithm using data from peripheral pulselessness and free-living conditions. Once developed, we evaluated the end-to-end algorithm prospectively: there was 1 unintentional emergency call per 21.67 user-years across two prospective studies; the sensitivity was 67.23% (95% confidence interval, 64.32%–70.05%) in a prospective arterial occlusion cardiac arrest simulation model. These results suggest a new opportunity, deployable at scale, for wearable-based detection of sudden loss of pulse while minimizing societal costs of excess false detections⁷.

Background Much of the early data on COVID-19 symptomatology was captured in the hospital setting. In a community setting the symptoms most predictive of SARS-CoV-2 positivity may be different. Data from the California sites of a COVID-19 community testing program are presented here. Methods Prior to being tested, participants in the Baseline COVID-19 Testing Program completed an online screener, in which they self-reported basic demographics and the presence or absence of 10 symptoms. Both positive and negative COVID-19 RT-PCR tests were linked back to the screener data. A multivariable model of positivity was fit using generalized estimating equations, adjusting for month of testing as a fixed effect and accounting for clustering of data within each test site. Results Among 547,018 first-time tests in California in 2020, positivity rates were 3.4%, 9.9%, and 19.8% for participants with no symptoms, 1 symptom, or 2 or more symptoms at the time of screening, respectively. All ten symptoms were individually associated with higher positivity rates, but only six of ten symptoms were associated with higher positivity when adjusting for other symptoms. Major symptoms with highest predictive value were recent loss of taste or smell, fever, and coughing with ORs of 3.27, 1.97, and 1.95, respectively. Shortness of breath and vomiting or diarrhea were negatively associated with positivity adjusting for other symptoms and, absent other symptoms, participants with these symptoms did not have significantly higher positivity rates than asymptomatic participants. Conclusions Recent loss of taste and smell should be elevated to a major symptom along with fever and coughing in public health messaging and in our community approach to testing and surveillance, while mild to moderate shortness of breath should be de-emphasized as a sensitive early predictor of COVID-19 positivity.

Background: The Onduo virtual diabetes clinic (VDC) for people with type 2 diabetes (T2D) combines a mobile app, remote personalized lifestyle coaching, connected devices, and live video consultations with board-certified endocrinologists for medication management and prescription of real-time continuous glucose monitoring (RT-CGM) devices for intermittent use. Objective: This prospective single-arm study evaluated glycemic outcomes associated with participation in the Onduo VDC for 4 months. Methods: Adults aged ≥18 years with T2D and a baseline glycated hemoglobin (HbA1c) of ≥8% to ≤12% were enrolled from 2 primary care centers from February 2019 to October 2019. Participants were asked to engage at ≥1 time per week with their care team and to participate in a telemedicine consultation with a clinic endocrinologist for diabetes medication review. Participants were asked to use a RT-CGM device and wear six 10-day sensors (total 60 days of sensor wear) intermittently over the course of 4 months. The primary outcome was change in HbA1c at 4 months from baseline. Other endpoints included change in weight and in RT-CGM glycemic metrics, including percent time 250 mg/dL. Changes in blood pressure and serum lipids at 4 months were also evaluated. Results: Participants (n=55) were 57.3 (SD 11.6) years of age, body mass index 33.7 (SD 7.2), and 40% (22/55) female. HbA1c decreased significantly by 1.6% (SD 1%; P<.001). When stratified by baseline HbA1c of 8.0% to 9.0% (n=36) and >9.0% (n=19), HbA1c decreased by 1.2% (SD 0.6%; P<.001) and 2.4% (SD 1.3%; P<.001), respectively. Continuous glucose monitoring–measured (n=43) percent time in range (TIR) 70-180 mg/dL increased by 10.2% (SD 20.5%; P=.002), from 65.4% (SD 23.2%) to 75.5% (SD 22.7%), which was equivalent to a mean increase of 2.4 hours TIR per day. Percent time 181-250 mg/dL and >250 mg/dL decreased by 7.2% (SD 15.4; P=.005) and 3.0% (SD 9.4; P=.01), respectively. There was no change in percent time <70 mg/dL. Mean weight decreased by 9.0 lb (SD 10.4; P<.001). Significant improvements were also observed in systolic blood pressure, total cholesterol, low-density lipoprotein cholesterol, and triglycerides (P=.04 to P=<.001). Conclusions: Participants in the Onduo VDC experienced significant improvement in HbA1c, increased TIR, decreased time in hyperglycemia, and no increase in hypoglycemia at 4 months. Improvements in other metabolic health parameters including weight and blood pressure were also observed. In conclusion, the Onduo VDC has potential to support people with T2D and their clinicians between office visits by increasing access to specialty care and advanced diabetes technology including RT-CGM. Trial Registration: ClinicalTrials.gov NCT03865381; https://clinicaltrials.gov/ct2/show/NCT03865381

BACKGROUND The Onduo virtual diabetes clinic (VDC) for people with type 2 diabetes (T2D) combines a mobile app, remote personalized lifestyle coaching, connected devices, and live video consultations with board-certified endocrinologists for medication management and prescription of real-time continuous glucose monitoring (RT-CGM) devices for intermittent use. OBJECTIVE This prospective single-arm study evaluated glycemic outcomes associated with participation in the Onduo VDC for 4 months. METHODS Adults aged ≥18 years with T2D and a baseline glycated hemoglobin (HbA1c) of ≥8% to ≤12% were enrolled from 2 primary care centers from February 2019 to October 2019. Participants were asked to engage at ≥1 time per week with their care team and to participate in a telemedicine consultation with a clinic endocrinologist for diabetes medication review. Participants were asked to use a RT-CGM device and wear six 10-day sensors (total 60 days of sensor wear) intermittently over the course of 4 months. The primary outcome was change in HbA1c at 4 months from baseline. Other endpoints included change in weight and in RT-CGM glycemic metrics, including percent time 250 mg/dL. Changes in blood pressure and serum lipids at 4 months were also evaluated. RESULTS Participants (n=55) were 57.3 (SD 11.6) years of age, body mass index 33.7 (SD 7.2), and 40% (22/55) female. HbA1c decreased significantly by 1.6% (SD 1%; P<.001). When stratified by baseline HbA1c of 8.0% to 9.0% (n=36) and >9.0% (n=19), HbA1c decreased by 1.2% (SD 0.6%; P<.001) and 2.4% (SD 1.3%; P<.001), respectively. Continuous glucose monitoring–measured (n=43) percent time in range (TIR) 70-180 mg/dL increased by 10.2% (SD 20.5%; P=.002), from 65.4% (SD 23.2%) to 75.5% (SD 22.7%), which was equivalent to a mean increase of 2.4 hours TIR per day. Percent time 181-250 mg/dL and >250 mg/dL decreased by 7.2% (SD 15.4; P=.005) and 3.0% (SD 9.4; P=.01), respectively. There was no change in percent time <70 mg/dL. Mean weight decreased by 9.0 lb (SD 10.4; P<.001). Significant improvements were also observed in systolic blood pressure, total cholesterol, low-density lipoprotein cholesterol, and triglycerides (P=.04 to P=<.001). CONCLUSIONS Participants in the Onduo VDC experienced significant improvement in HbA1c, increased TIR, decreased time in hyperglycemia, and no increase in hypoglycemia at 4 months. Improvements in other metabolic health parameters including weight and blood pressure were also observed. In conclusion, the Onduo VDC has potential to support people with T2D and their clinicians between office visits by increasing access to specialty care and advanced diabetes technology including RT-CGM. CLINICALTRIAL ClinicalTrials.gov NCT03865381; https://clinicaltrials.gov/ct2/show/NCT03865381

In the original version of the published article, line three of the third paragraph of the methods stated “Excluding patients with micrometastatic disease, the 5-year BCSM for patients with Recurrence Score results n = 2,617) was 1.3% (95% CI, 0.6–2.9%).” To improve clarity this statement has been replaced with “Excluding patients with micrometastatic disease, there were 2,617 patients with 1–3 positive nodes. Of these, 1,487 also had Recurrence Score results <18 with 5-year BCSM of 1.3% (95% CI, 0.6–2.9%).” The original version of the published article also contained an error in the second sentence of the Figure 2 legend describing the mutation status of the patient population examined. The sentence in the original published version of the article stated “Patients with HR+, HER2-positive, node-negative…” this has been changed to “Patients with HR+, HER2-negative, node-negative…”. This has been corrected in the PDF and HTML versions of this paper.

Purpose: Limited data exist on the molecular biology, treatment, and outcomes of breast cancer in men, and much of our understanding in this area remains largely an extrapolation from data in women with breast cancer. Materials and Methods: We studied men and women with hormone receptor–positive breast cancer and the 21-gene Breast Recurrence Score (RS) results. Differences in clinical characteristics and gene expression were determined, and distribution of RS results was correlated with 5-year breast cancer–specific survival (BCSS) and overall survival. Results: There were 3,806 men and 571,115 women. Men were older than women (mean age, 64.2 v 59.1 years; P < .001). RS < 18 predominated in both genders, but RS ≥ 31 was more frequent in men (12.4% v 7.4%; P < .001), as were very low scores (RS < 11; 33.8% v 22.1%; P < .001). Mean gene expression was higher in men for the estrogen receptor (ER), proliferation, and invasion groups. ER was lowest and progesterone receptor was highest in women younger than 50 years of age, with a progressive increase in ER with age. Men younger than 50 years of age had slightly lower ER and progesterone receptor compared with older men. Survival data were available from SEER for 322 men and 55,842 women. Five-year BCSS was 99.0% (95% CI, 99.3% to 99.9%) and 95.9% (95% CI, 87.6% to 98.7%) for men with RS < 18 and RS 18-30, respectively, and for women, it was 99.5% (95% CI, 99.4% to 99.6%) and 98.6% (95% CI, 98.4% to 98.8%), respectively. RS ≥ 31 was associated with an 81.0% 5-year BCSS in men (95% CI, 53.3% to 93.2%) and 94.9% 5-year BCSS (95% CI, 93.9% to 95.7%) in women. Five-year BCSS and overall survival were lower in men than in women. Conclusion: This study reveals some distinctive biologic features of breast cancer in men and an important prognostic role for RS testing in both men and women.

11568 Background: Linking the 21-gene assay RS result to the SEER Registries demonstrated very low 5-y BCSM with low RS and high 5-y BCSM with high RS across subgroups, such as nodal status, age, tumor size and grade (npj Breast Cancer 2016). Given the large sample size and interest in outcomes as a function of tumor characteristics, we characterized the relationship between RS results and BCSM in patients reported by SEER with lobular morphology. Methods: Patients with RS and lobular morphology based on the registry ICD-O-3 code 8520 were eligible if node negative (N0) or node positive up to 3 positive nodes (N+mic,1-3), HR+, HER2- negative, no prior malignancy, and diagnosed between Jan 2004 and Dec 2012. No information in SEER is available regarding lobulars, ie., trabecular, alveolar, solid and pleomorphic. 5-y BCSM was estimated using actuarial methods. Results: There were 6,075 eligible patients reported with lobular morphology (11% of cases). Median age was 59 years; 88%/12% were N0/N+; 31%/62%/7% grade 1/2/3; 61%/39% ≤2 cm/>2 cm. Median follow-up was 44 months. A minority (8%) had RS >25. Chemotherapy (CT) use and BCSM increased with increasing RS. In multivariable analysis in N0 disease, continuous RS result and tumor size predicted BCSM (p=0.003 and p=0.04, respectively), whereas age and tumor grade were non-significant. In multivariable analysis in N+ disease, continuous RS result alone predicted BCSM (p=0.002). Conclusions: In these analyses the prognosis of patients with lobular breast cancer treated based on RS results depends on both nodal status and the RS result. The 5-y BCSM for lobular breast cancer is excellent with RS of 25 or less, and increases for RS >25. [Table: see text]

537 Background: The Recurrence Score (RS) was shown in NSABP B-20 to predict chemotherapy (CT) benefit for RS ≥31 and no CT benefit for RS <18. The TAILORx results for RS <11 (NEJM 2015) reported excellent outcomes with no opportunity for CT to add additional benefit. As we await TAILORx results for RS 11-25, we characterized BC specific mortality (BCSM) for RS groups (cutoffs of 11, 18, 25, and 31) in the population-based SEER study of pts treated based on RS. Methods: RS results were provided to SEER registries per their methods (npj Breast Cancer 2016). Pts diagnosed (Jan 2004 - Dec 2012) with N0 HR+ HER2- negative BC, and no prior malignancy were eligible. BCSM estimates by CT use were computed using standard cutpoints of 18 and 31 and TAILORx cutpoints of 11 and 25. Results: Among 49,681 with a RS, 9,486 (19%) had RS <11, 17,988 (36%) had RS 11-17, 14,541 (29%) had RS 18-25, 3,805 (8%) had RS 26-30, and 3,861 (8%) had RS ≥31. Reported CT use and 5-y BCSM increased with increasing RS. For pts with both RS <11 and RS 11-17, CT use was uncommon and 5-y BCSM was low regardless of CT use. For pts with RS 18-25, CT use was more common and the 5-y BCSM was about 1% regardless of CT use. For pts with RS of 26-30 or ≥31, CT was common, and lower 5-y BCSM was observed with CT reported yes than with CT reported no or unknown. Conclusions: Pts in real-world clinical practice with RS <11, consistent with TAILORx, and pts with RS 11-17 have low 5-y BCSM with limited CT use, supporting hormonal therapy alone for pts with RS <18. The high end of the TAILORx mid-range (18-25) also showed good 5-y BCSM both with and without CT, highlighting the importance of the randomized results of TAILORx.

528 Background: Better tools are needed to estimate the risk of local recurrence (LR; DCIS or invasive) after breast-conserving surgery (BCS) for DCIS to inform treatment decisions. The DCIS Score (DS) was validated as a predictor of LR in E5194 and Ontario DCIS Cohort (ODC) after BCS without radiation (Solin,2013; Rakovitch,2015). We performed a meta-analysis (MA) combining data from E5194 and ODC with additional follow-up from E5194 adjusting for pertinent clinico-pathologic factors to provide refined prediction estimates of LR risk after BCS alone. Methods: The MA used data from E5194 and ODC. Patients with positive margins and multifocality were excluded. Identical Cox regression models were fit including age at diagnosis ( < 50, ≥50 yr), tumor size (1cm, > 1cm), DCIS Score and year of surgery (before vs after 2000). Grade was not significant. MA was used to calculate precision-weighted estimates of 10 year LR risk by DS. Results: Combined cohort includes 773 pts (tamoxifen used in 20% E5194, 17% of ODC > 65 yr). The DS and the clinico-pathologic variables age, tumor size and year provided independent prognostic information on 10 yr LR risk (p≤.009). Hazard ratios from E5194 and ODC cohorts were similar for tumor size ≤1 vs. > 1cm (1.45, 1.47), age ≥50 vs. < 50 yr (0.61, 0.84) and surgery year after 2000 (0.67, 0.49). 10 yr LR risks by combinations of age, tumor size, and DS are detailed in Table. For patients ≥50 yr with tumors ≤1cm and low risk DS, the 10 yr LR risks range from 5.3-10.0%. A high risk DS is associated with a higher 10 yr predicted risk of LR in all subsets. 10 yr risk of contralateral BC was 5.4%. Conclusions: This MA provides refined estimates of 10 yr LR risk after BCS alone for DCIS. Adding clinico-pathologic factors to the DCIS Score provides enhanced prognostic LR risk estimates to guide individualized treatment decision-making.