The analysis used an individual-level state-transition microsimulation model built with monthly cycles and a lifetime horizon. The simulated cohort comprised female patients aged 49 years newly diagnosed with operable stage II or III triple-negative breast cancer (TNBC). The model incorporated the currently recommended treatment and surveillance pathways to represent standard clinical practice.
Model inputs followed the National Comprehensive Cancer Network guidance for surveillance after initial diagnosis and curative-intent therapy. Treatment and follow-up modalities recommended in current guidelines were integrated into the simulation so that surveillance events and potential recurrence detection reflected standard-of-care practice.
Outcomes assessed in the model included cumulative recurrences, detected recurrences, detection rate among recurrences, overall survival (OS) and recurrence-free survival (RFS) over a 5-year horizon, and average life expectancy. The analysis focused on the effectiveness of guideline-based surveillance in identifying recurrences and reported resource use metrics linked to detection events.
The model estimated a 5-year overall survival rate of 81.4% ± 0.1% (SE) for the simulated cohort of patients with stage II or III TNBC. The 5-year recurrence-free survival was 78.3% ± 0.1% (SE).
Within the same 5-year period, the model projected that 7.4% ± 0.1% (SE) of patients would experience recurrence. Of those recurrences, the breakdown by site was 68% distant and 32% locoregional.
Despite guideline-concordant surveillance, the model found a relatively low detection rate: only 59.3% ± 0.2% (SE) of modeled recurrences were detected by the recommended follow-up processes. This implies that a substantial proportion of recurrences were not identified through the current standard surveillance strategy within the 5-year window.
The model provided resource-efficiency estimates tied to detection yield. It projected that 399 mammograms would be required to detect a single locoregional recurrence. For distant recurrences, the estimate was 279 other imaging tests required to identify one distant recurrence. These figures quantify the high volume of imaging needed per recurrence detected under existing surveillance recommendations.
The simulation highlights limitations in the effectiveness of current guideline-based surveillance for detecting recurrences in operable stage II and III TNBC. With fewer than two-thirds of recurrences identified and large numbers of imaging tests needed per detected event, the model points to gaps between recommended follow-up and timely recurrence detection in this high-risk subtype.
The authors emphasize that improved surveillance approaches are needed during the follow-up period for patients with TNBC. The clinically validated microsimulation can be used to explore how alternative strategies or novel testing modalities might change detection rates, resource utilization, and ultimately patient outcomes.
The model was described as clinically validated and designed to reflect current practice. Because it integrates guideline-directed treatment and surveillance and produces quantitative outputs for detection and survival, it can serve as a platform to evaluate emerging surveillance technologies or changes to follow-up protocols. The source notes that as new testing methods become available, this model can be applied to assess whether they improve recurrence detection and patient outcomes.
The source article presents model results and their implications but does not report every parameter or external validation metric in the summary provided here. Details not specified in the source text — such as the full list of model input parameters, calibration targets beyond the reported outcomes, or sensitivity analysis results — were not reported in the provided summary.
In this microsimulation of patients with operable stage II and III TNBC managed according to current guideline surveillance, 5-year survival estimates were favorable but recurrence detection was limited. A detection rate of 59.3% and high imaging-per-detection estimates (399 mammograms per locoregional recurrence; 279 other imaging tests per distant recurrence) underscore the need for improved surveillance strategies. The model offers a tool to compare novel testing approaches against the existing standard of care.