Increasing knowledge about signal transduction pathways as drivers of cancer growth has elicited the development of “targeted drugs,” which inhibit aberrant signaling pathways. They require a companion diagnostic test that identifies the tumor-driving pathway; however, currently available tests like estrogen receptor (ER) protein expression for hormonal treatment of breast cancer do not reliably predict therapy response, at least in part because they do not adequately assess functional pathway activity. We describe a novel approach to predict signaling pathway activity based on knowledge-based Bayesian computational models, which interpret quantitative transcriptome data as the functional output of an active signaling pathway, by using expression levels of transcriptional target genes. Following calibration on only a small number of cell lines or cohorts of patient data, they provide a reliable assessment of signaling pathway activity in tumors of different tissue origin. As proof of principle, models for the canonical Wnt and ER pathways are presented, including initial clinical validation on independent datasets from various cancer types. Cancer Res; 74(11); 2936–45. ©2014 AACR.

The idea of artificial intelligence (AI) has a long history. It turned out, however, that reaching intelligence at human levels is more complicated than originally anticipated. Currently, we are experiencing a renewed interest in AI, fueled by an enormous increase in computing power and an even larger increase in data, in combination with improved AI technologies like deep learning. Healthcare is considered the next domain to be revolutionized by artificial intelligence. While AI approaches are excellently suited to develop certain algorithms, for biomedical applications there are specific challenges. We propose six recommendations—the 6Rs—to improve AI projects in the biomedical space, especially clinical health care, and to facilitate communication between AI scientists and medical doctors: (1) Relevant and well-defined clinical question first; (2) Right data (ie, representative and of good quality); (3) Ratio between number of patients and their variables should fit the AI method; (4) Relationship between data and ground truth should be as direct and causal as possible; (5) Regulatory ready; enabling validation; and (6) Right AI method.

Endocrine therapy is important for management of patients with estrogen receptor (ER)–positive breast cancer; however, positive ER staining does not reliably predict therapy response. We assessed the potential to improve prediction of response to endocrine treatment of a novel test that quantifies functional ER pathway activity from mRNA levels of ER pathway–specific target genes. ER pathway activity was assessed on datasets from three neoadjuvant-treated ER-positive breast cancer patient cohorts: Edinburgh: 3-month letrozole, 55 pre-/2-week/posttreatment matched samples; TEAM IIa: 3- to 6-month exemestane, 49 pre-/28 posttreatment paired samples; and NEWEST: 16-week fulvestrant, 39 pretreatment samples. ER target gene mRNA levels were measured in fresh-frozen tissue (Edinburgh, NEWEST) with Affymetrix microarrays, and in formalin-fixed paraffin-embedded samples (TEAM IIa) with qRT-PCR. Approximately one third of ER-positive patients had a functionally inactive ER pathway activity score (ERPAS), which was associated with a nonresponding status. Quantitative ERPAS decreased significantly upon therapy (P < 0.001 Edinburgh and TEAM IIa). Responders had a higher pretreatment ERPAS and a larger 2-week decrease in activity (P = 0.02 Edinburgh). Progressive disease was associated with low baseline ERPAS (P = 0.03 TEAM IIa; P = 0.02 NEWEST), which did not decrease further during treatment (P = 0.003 TEAM IIa). In contrast, the staining-based ER Allred score was not significantly associated with therapy response (P = 0.2). The ERPAS identified a subgroup of ER-positive patients with a functionally inactive ER pathway associated with primary endocrine resistance. Results confirm the potential of measuring functional ER pathway activity to improve prediction of response and resistance to endocrine therapy.

The shift toward remote patient monitoring methods to detect clinical deterioration requires testing of wearable devices in real-life clinical settings. This study aimed to develop a remote early warning scoring (REWS) system based on continuous measurements using a wearable device, and compare its diagnostic performance for the detection of deterioration to the diagnostic performance of the conventional modified early warning score (MEWS). The study population of this prospective, single center trial consisted of patients who underwent major abdominal cancer surgery and were monitored using routine in-hospital spotcheck measurements of the vital parameters. Heart and respiratory rates were measured continuously using a wireless accelerometer patch (HealthDot). The prediction by MEWS of deterioration toward a complication graded Clavien-Dindo of 2 or higher was compared to the REWS derived from continuous measurements by the wearable patch. A total of 103 patients and 1909 spot-check measurements were included in the analysis. Postoperative deterioration was observed in 29 patients. For both EWS systems, the sensitivity (MEWS: 0.20 95% CI: [0.13-0.29], REWS: 0.20 95% CI: [0.13-0.29]) and specificity (MEWS: 0.96 95% CI: [0.95-0.97], REWS: 0.96 95% CI: [0.95-0.97]) were assessed. The diagnostic value of the REWS method, based on continuous measurements of the heart and respiratory rates, is comparable to that of the MEWS in patients following major abdominal cancer surgery. The wearable patch could detect the same amount of deteriorations, without requiring manual spot check measurements.