Why This Study Matters
Metastatic breast cancer remains incurable, and patients often endure prolonged, sequential systemic therapies during which disease burden and treatment toxicity accumulate. Fatigue — a persistent sense of physical, emotional, and cognitive exhaustion — is among the most common and debilitating symptoms in this population, yet it is frequently underrecognized in routine care. Electronic patient-reported outcome (ePRO) monitoring has shown promise for improving symptom control in advanced cancer broadly, but disease-specific, multicenter randomized evidence in metastatic breast cancer has been limited — a gap the PRO B trial was designed to close.
Study Design
PRO B was an investigator-initiated, open-label, multicenter randomized clinical trial conducted at 52 certified academic and community breast cancer centers across Germany. Patients were randomly assigned 1:1, stratified by metastatic site distribution, histological subtype, and study center, to weekly alert-based ePRO monitoring or usual quarterly PRO monitoring. Of 2,008 patients screened, 924 were randomized, with 909 making up the modified intention-to-treat population.
| Element | Detail |
|---|---|
| Design | Investigator-initiated, open-label, multicenter randomized clinical trial |
| Randomization | 1:1, stratified by metastatic site distribution, histological subtype, and study center |
| Intervention arm (n=456 analyzed) | Weekly ePRO questionnaires via smartphone app (EORTC QLQ-C30 CAT core item bank); clinically meaningful deterioration triggered clinician telephone contact within 48 hours |
| Control arm (n=453 analyzed) | Usual care with the full PRO questionnaire completed quarterly; no alerts generated, data not visible to the care team |
| Primary endpoint | Fatigue (EORTC QLQ-C30 CAT core item bank T score) at 6 months |
| Study sites | 52 academic and community breast cancer centers, Germany |
Patient Population
Eligible patients were women 18 years or older with metastatic breast cancer receiving systemic therapy, with a life expectancy longer than 3 months, ECOG performance status of 0-2, German literacy, and smartphone access. Mean age was approximately 51 years in both groups (range 19-83). Hormone receptor-positive, ERBB2-negative disease was the most common subtype (66.2% of patients), and 41.8% had de novo metastatic presentation at diagnosis. Baseline characteristics were generally balanced between groups, though baseline PRO scores — including fatigue — favored the intervention group in most domains, a limitation the authors flag explicitly in their discussion.
Primary Endpoint Results
The trial met its primary endpoint. At 6 months, adjusted mean fatigue was 54.5 (95% CI, 53.7-55.4) in the weekly-monitoring group versus 59.9 (95% CI, 59.0-60.8) in the usual-care group — a mean difference of -5.4 points (95% CI, -6.6 to -4.1; P<.001). This exceeded the prespecified minimal clinically important difference of 3.3 points, and sensitivity analyses across different missing-data assumptions and analytic approaches showed the between-group difference remained consistent in direction and magnitude.
Secondary patient-reported outcomes over 12 months also favored weekly monitoring: fatigue differences were -5.2 points at 3 months, -6.8 points at 9 months, and -5.2 points at 12 months. Physical functioning differences were 3.7 points at 3 months, 4.0 points at 6 months, and 5.1 points at 9 months.
Exploratory clinical outcomes were mixed. Time to first change in systemic therapy did not differ between groups (adjusted hazard ratio 0.89). In an additional exploratory analysis, full-cohort 12-month overall survival was 87.6% in the weekly-monitoring group versus 84.7% in the usual-care group (adjusted hazard ratio 0.72). The authors are explicit that this survival signal is exploratory and hypothesis-generating rather than confirmatory, since the trial was not powered for survival and loss to follow-up was higher in the control arm.
Subgroup Analyses
Descriptive subgroup estimates for the primary fatigue endpoint consistently favored weekly monitoring across every stratum tested, and no treatment-by-subgroup interaction reached statistical significance.
| Subgroup | Intervention T Score | Control T Score | Mean Difference (95% CI) |
|---|---|---|---|
| Age <50 years | 54.4 | 60.4 | -6.0 (-7.9 to -4.1) |
| Age >=50 years | 54.7 | 59.6 | -4.9 (-6.5 to -3.2) |
| Visceral metastases (single organ) | 54.4 | 61.8 | -7.5 (-10.8 to -4.1) |
| Triple-negative breast cancer | 54.3 | 61.6 | -7.3 (-11.2 to -3.3) |
| HR-positive/ERBB2-positive | 53.9 | 59.1 | -5.1 (-8.2 to -2.1) |
| HR-negative/ERBB2-positive | 52.9 | 61.5 | -8.5 (-13.6 to -3.5) |
Protocol-specified exploratory survival subgroup analyses were reported for two prespecified populations: patients with visceral metastases (adjusted hazard ratio 0.37) and patients with triple-negative breast cancer (adjusted hazard ratio 0.73). Among patients with hormone receptor-positive, ERBB2-negative disease, 12-month survival was 88.5% versus 82.7% (adjusted hazard ratio 0.69). The authors describe all survival subgroup findings as exploratory and underpowered.
Safety Profile
Because this trial evaluated a digital monitoring intervention rather than a pharmacologic agent, no conventional drug-related adverse-event table was reported. The closest analog is the pattern of alert-triggered clinical contacts and their outcomes.
| Alert Outcome | Count (%) |
|---|---|
| Total alerts generated (of 24,180 follow-up assessments) | 3,019 (12.5%) |
| Led to patient contact | 2,559 |
| Contact within 48 hours | 2,194 (85.7%) |
| No intervention/counseling required | 1,241 (48.5%) |
| Resolved by telephone consultation | 1,131 (44.2%) |
| Occurred while already hospitalized | 109 (4.3%) |
The most common documented reasons for telephone consultation were infection or colds (14.6%), therapy-related adverse effects (8.1%), and musculoskeletal pain (5.8%). In an exploratory claims-data subset covering about a quarter of the cohort (235 of 909 patients), 12-month hospitalization and emergency department visit rates were lower with weekly monitoring (incidence rate ratios of 0.52 and 0.76, respectively) — though the authors caution that claims data covered only patients insured by three participating statutory health insurers, so these results should be read as supportive rather than definitive.
In this multicenter randomized clinical trial, alert-based PRO monitoring was associated with clinically meaningful and sustained reductions in fatigue and improvements in PF in patients with mBC. However, the results should be interpreted cautiously.
— Discussion, PRO B trial (Karsten et al., JAMA Oncology, 2026)
Interpretation and Broader Context
The authors identify several important caveats. Because the trial was open-label and baseline PRO assessments were collected after randomization, participants may have become aware of their allocation, potentially introducing expectancy effects or contributing to the baseline imbalance in fatigue scores. Questionnaire completion rates were higher in the intervention group, and attrition was greater among patients with poorer self-reported performance status — both of which could have amplified the observed between-group differences, particularly at later timepoints. The primary endpoint was also changed early in the trial, from 12-month to 6-month fatigue, following an early protocol amendment.
The reported survival benefit and reduced health care utilization were both explicitly labeled exploratory and hypothesis-generating by the study authors. The trial was not powered to detect survival differences, and health care utilization data covered only about one-quarter of the cohort. The authors conclude that alert-based ePRO monitoring may improve symptom-focused care for patients with metastatic breast cancer, and that its use across certified breast cancer centers supports relevance to routine oncology workflows — particularly where digital PRO monitoring is not yet standard — but that effects on longer-term outcomes, health care utilization, and survival require confirmation in future studies.
In the phase 3 PRO B trial, weekly alert-based electronic patient-reported outcome monitoring produced a clinically meaningful and durable reduction in fatigue (mean difference -5.4 points at 6 months, sustained through 12 months) and improved physical functioning compared with usual quarterly monitoring, in patients with metastatic breast cancer. Exploratory signals for reduced hospitalization and improved survival are promising but unconfirmed, and the open-label design with baseline imbalances means these findings should motivate broader adoption of digital symptom monitoring in metastatic breast cancer care while awaiting further confirmatory evidence.