Challenging the cornerstone: Toward risk-adapted follow-up in early breast cancer

The follow up in early breast cancer represented a major challenging in oncology practice. For decades, routine radiological and laboratory surveillance has been widely adopted and often perceived as a reassuring and protective strategy for both patients and clinicians. This approach has persisted despite evidence from randomized clinical trials demonstrating no survival benefit from routine imaging or tumour marker assessment in asymptomatic patients.^1–4

This observation highlights a crucial gap: the evidence demonstrating lack of benefit from intensive follow-up largely predates the era of targeted therapies, immunotherapy, advanced imaging, and biomarker-guided management, and was not designed to test risk-adapted surveillance strategies consistent with current standards of care. ⁵

The primary aim of this paper is not to advocate an immediate change in current clinical practice, nor simply to revisit the historical evidence showing the lack of benefit of intensive surveillance. Rather, this article seeks to frame the issue of follow-up in early breast cancer within the context of contemporary oncology and real-world healthcare systems. This perspective aims to stimulate a more nuanced discussion on how surveillance strategies should evolve in an era of biologically stratified breast cancer care.

A Cochrane review published in 2016 reported no improvement in survival or disease-related outcomes across five multicentre randomized controlled trials comparing intensive follow-up strategies including scheduled imaging and laboratory testing (such as chest radiography, liver ultrasound or CT scans, bone scans, and tumour marker assessment) with less intensive, symptom-driven follow-up based on clinical history and physical examination in women treated for early breast cancer, and estimated that approximately 87,670 follow-up visits were needed for healthcare providers to detect a single curable recurrence.^6,7 The Kronos trial compared standard follow up (yearly mammography and physical exam) with an intensified approach (standard plus quarterly serum CEA and CA 15-3, with PET-FdG18 scan performed if markers rise critically). The primary end point was overall survival. No differences emerged between the two cohorts even if potentially critical diagnostic anticipation has been observed in the experimental arm in the Her 2 positive population. ⁸ Similarly, an Italian observational study showed that an intensive survivorship care plan with tumour markers and/or routine imaging tests did not improve overall survival. ⁹

Despite strong evidence against routine intensive follow-up, many oncologists continue to implement such schedules in practice, highlighting a persistent gap between guideline recommendations and real-world care. Moreover, the benefit of routine surveillance is increasingly questioned considering the heterogeneous recurrence patterns across breast cancer subtypes, underscoring the need for risk-adapted, biologically informed follow-up strategies. In modern breast cancer management, prognosis and treatment decisions are informed not only by anatomical stage (tumour size and nodal involvement) but also by tumour biology, defined by intrinsic subtypes (luminal A and B, HER2-positive, and triple-negative).^4,10

These subtypes exhibit distinct natural histories and recurrence patterns and directly guide systemic treatment choices, including neoadjuvant and adjuvant chemotherapy, endocrine therapy, anti-HER2 agents, and other targeted treatments.^11,12

Applying a uniform follow-up strategy in an era of increasingly stratified and personalized therapy is therefore conceptually inconsistent. Large long-term cohorts further illustrate this biological heterogeneity. The annualized risk of recurrence is highest during the first five years after diagnosis (10.4%), with a peak between the first and second year (15.2%). ¹³

During this early phase, patients with oestrogen receptor–positive disease experience a lower risk compared with ER-negative disease (9.9% vs 11.5%; P = 0.01). Beyond five years, however, ER-positive disease is characterized by a sustained long-term risk of recurrence that exceeds that of ER-negative disease over extended follow-up, persisting for up to two decades. These time- and subtype-dependent patterns cannot be adequately captured by ‘one-size-fits-all’ follow-up algorithms. Luminal cancers may recur many years even decades after initial treatment, making lifelong intensive surveillance impractical, whereas aggressive subtypes such as triple-negative often relapse rapidly, sometimes during (neo)adjuvant therapy. HER2-positive disease represents a dynamic setting, as the advent of highly effective targeted therapies is reshaping its natural history and risk profile, significantly reducing the risk of recurrence and substantially improving outcomes even when the disease relapses at a later stage.

Rather than advocating indiscriminate intensification of surveillance, a more rational and biologically coherent approach is the adoption of risk-adapted follow-up strategies, conceptually aligned with the escalation and de-escalation principles that already underpin modern adjuvant systemic therapy. Within this framework, the intensity and duration of follow-up are tailored according to baseline risk, intrinsic subtype, response to treatment, and residual disease burden, while also addressing treatment-related toxicities and broader survivorship needs.

Emerging tools—including circulating tumour DNA (ctDNA) monitoring, novel biomarkers, and multivariable prognostic models—hold promise for improving the identification of patients at increased risk of locoregional or distant recurrence. Recent studies suggest that ctDNA analysis may allow earlier detection of molecular residual disease and impending relapse.

The Survive trial (NCT05658172) a double-blinded superiority trial is currently ongoing. The study compares intensified liquid biopsy-guided surveillance with standard follow-up in medium-to-high-risk early breast cancer patients. If biomarkers become positive or symptoms appear, staging exams are performed. A total of 3500 patients will be randomized equally after completing primary treatment. The main endpoints are overall survival and lead time effect. ¹⁴

However, at present, the clinical utility of these approaches remains uncertain. Optimal thresholds, timing, and integration with imaging modalities have not yet been defined, and their incorporation into routine follow-up practice requires rigorous prospective validation in cohorts treated according to contemporary therapeutic standards.

Taken together, these considerations underscore the need for modern, prospective studies specifically designed to evaluate biologically informed, risk-adapted follow-up strategies. Such studies should aim to identify patient subgroups who may—or may not—derive clinical benefit from intensified surveillance, rather than continuing to apply uniform follow-up protocols to all breast cancer survivors.

A shift toward personalized, evidence-based follow-up—considering disease burden, tumour subtype, treatment response, and, potentially, ctDNA dynamics—should be pursued, adopting principles of escalation and de-escalation already established in adjuvant treatment decision-making. Consequently, diversified, risk-adapted surveillance strategies should be developed, with intensified follow-up in higher-risk patients and de-escalated approaches in those at lower risk.

This strategy has the potential to reduce unnecessary interventions and associated harms, enhance informed patient participation, and promote ethical and sustainable use of healthcare resources. Nevertheless, the implementation of risk-based follow-up requires robust clinical evidence, which is currently limited. Predictive models, novel diagnostic tools, and biomarker assessments may play a crucial role in identifying patients at higher risk of locoregional or distant recurrence, but the optimal biomarkers, imaging modalities, and timing of assessments remain to be defined through well-designed prospective studies. In addition, the potential harms associated with intensive follow-up strategies—including cumulative radiation exposure and false-positive findings—should not be overlooked, as they may lead to psychological distress, anxiety, and a consequent deterioration in patients’ quality of life (QoL). Future prospective studies should therefore incorporate appropriate tools for longitudinal patient monitoring; in this context, Patient-Reported Outcomes (PROs) and QoL assessments may provide valuable insights into the long-term physical, psychological, and social consequences of cancer diagnosis and treatment. ¹⁵

Beyond surveillance, reducing the risk of recurrence also requires greater investment in prevention and lifestyle interventions. Reframing follow-up as a calibrated, evidence-based clinical decision—rather than an automatic and uniform procedure represents both a scientific responsibility and a civic imperative in the era of personalized, sustainable healthcare.