Chronic Low-Grade Inflammation Precedes Early Locoregional Recurrence following Mastectomy in Low-Risk Early-Stage Breast Cancer: A Longitudinal Case Report

Abstract

Introduction

Early locoregional recurrence in low-risk early-stage breast cancer following mastectomy and adjuvant chemotherapy is a relatively infrequent event. Sufficient empirical evidence is sparse to support commonly accepted explanations relating to inflammation as a recurrence risk factor that precedes and likely promotes cancer recurrence evolution.

Case Presentation

Dynamics of circulating tumor cells, tissue change, and low-grade inflammation were recorded in a representative breast cancer survivor follow-up case having relapsed 31 months after treatment completion. Prospective longitudinal testing was conducted based on a cell-based liquid biopsy platform supporting multi-rare cell marker analysis. The study provided evidence for the inflammation theory relating to epithelial tissue inflammation detectable at baseline, steady progression in marker abnormality, and chronic low-grade systemic inflammation appearing in cyclical patterns.

Conclusion

Therefore, our observations of persistent inflammation prior to relapse provide deeper insights into pathology progression and contribute to emerging data to support the inflammation hypothesis.

Introduction

Despite advances in the effectiveness of breast cancer therapies over recent decades, a significant risk of relapse remains in patients assessed to be at low-risk, such as T1-2 N0 staged patients who have undergone mastectomy [1–3]. The identification of a subset of this unremarkable patient group that may benefit from post-mastectomy radiotherapy is of clinical interest. Locoregional recurrence (LRR) risk was determined by a wide spectrum of factors comprising age, tumor size, nodal status, grade, lymphovascular invasion, fascial involvement, pre-menopausal status, molecular subtypes, prior treatments [3, 4], as well as margin status [2]. Statistical evidence of an overall worse prognosis is further given by minimal residual disease (MRD) assessment based on the detection of disseminated tumor cells in bone marrow [5, 6] and by the presence of circulating tumor cells (CTCs) or circulating tumor DNA (ctDNA) in blood [7–9]. Recent progress in assessing relapse risk has been made by tracking patient-specific mutations using ctDNA liquid biopsy, achieving 100% specificity and 80% sensitivity. A positive ctDNA result reliably predicted relapse, with a median lead time of 11.7 months [10]. The high accuracy may be explained by the body’s inability to eliminate treatment-resistant clonogenic residual tumor cells, which persist as latent micro-lesions at the tumor margins and allegedly denote the source of ctDNA. These cells exploit, or may themselves establish, a favorable tumor microenvironment (TME), thereby contributing to a high probability of disease recurrence [11–13]. In analogy to the TME theory, infection-associated surgical site wound complications were also proposed as an inflammatory trigger to promote cancer recurrence evolution [13]. Population-level statistical evidence may be in favor of the dormancy/reactivation model to explain recurrence [14] and align with the idea of inflammation as a reactivation trigger. It is evident that, regardless of TME or infections, chronic inflammatory responses denote failure to heal and potentially accompany and drive cancer progression. The link between inflammation and cancer recurrence is well established [15, 16]. In this context, elevated pro-inflammatory factors prior to relapse have been proposed as predictors of increased recurrence risk [17]. Furthermore, the duration of inflammation and the balance between pro- and anti-inflammatory signals have been suggested to influence the course of recurrence [18].

Our data support and may further refine the inflammation model potentially applicable to breast cancer survivors with a low recurrence risk, having provided direct evidence of persistent systemic and local inflammation that is more cyclic in pattern. Residual cancer-related inflammation, irrespective of its origin, is proposed here to be detectable through multi-marker liquid biopsy, enabling simultaneous assessment of tissue damage, systemic cancer, and inflammatory burden [19, 20]. This platform builds on the systemic cytology principle [21] and harnesses the extensive diagnostic information derived from diverse circulating rare cells. This approach is proposed to provide improved consistency in detecting and characterizing disease progression. We investigated a representative low-risk HR+/HER2- T2N0 case in prospective longitudinal testing based on the multi-marker liquid biopsy platform with the initial indication of a residual cancer inflammation profile. To our knowledge, this is the first prospective longitudinal case report documenting the evolution of chronic low-grade inflammation and rare-cell abnormalities preceding LLR in a low-risk early-stage breast cancer survivor. The results demonstrated persistence and worsening of rare cell abnormalities over the entire period until relapse and a more complex inflammation process than commonly anticipated relating to inflammation cycles and nonlinear marker progression. The herein provided evidence supports the association of inflammation with cancer relapse and sets the ground for rare cell-based risk factor development ultimately aiming to improve prognostication in low-risk early-stage breast cancer survivors.

Case Presentation

Patient Information

The 67-year-old patient presented to a local hospital reporting breast pain and the discovery of a firm palpable mass. The patient had no prior history of cancer, hospitalizations, chronic medical conditions, or regular medication use and reported good overall health before diagnosis. No significant family history of cancer was reported, suggesting a low hereditary predisposition. The patient reported low levels of physical activity and was mildly obese (BMI 32 kg/m2). The patient remained asymptomatic until diagnosis and treatment, apart from breast pain prior to surgery. Mammography and core needle biopsy showed cancer in the palpable mass. The patient underwent mastectomy without breast reconstruction and sentinel node biopsy, which was subsequently confirmed as a 4.0 cm invasive lobular carcinoma of the breast (stage pT2N0M0), with ER positivity of 30%, PR positivity of 5%, and HER2-negative status. No cancer was found in 11 lymph nodes. Her adjuvant therapy included 6 cycles of cyclophosphamide, methotrexate, 5-fluorouracil regimen, and an aromatase inhibitor (anastrozole, 1 mg daily) as endocrine treatment. Adjuvant chemotherapy was initiated 1 month postoperatively and continued for 8 months. The patient was stratified as low risk; therefore, adjuvant chest wall radiation was not indicated. The patient did not report illness or mass palpation during the follow-up period until relapse. Serial blood-based follow-up monitoring commenced 2 weeks after completion of chemotherapy, denoting the individual baseline of rare cell abnormality and continued until the first detection of abnormal findings, extending the active monitoring period to 39 months postoperatively. Since the patient remained asymptomatic throughout the monitoring period, testing was not prompted by any clinical signs or subjective suspicion of disease progression. The blood sampling frequency depended on the subject’s availability for blood donation and resulted in a total of 7 tests during the follow-up period. Further circulating rare cell blood tests were conducted at 10, 12, 16, 19, 21, and then 31 months post-therapy completion. Based on the finding of rising rare cell abnormality (see results) in the blood 25 months after surgery, a bone scan and CT scan of the chest and upper abdomen were done, but did not show evidence of recurrence. A new nodule was found at the mastectomy site 39 months after surgery based on physical examination and ultrasound of the chest wall, which proved to be a recurrent invasive lobular carcinoma, of size 2.5 cm, with ER positive 70%; PR negative, HER-2 negative, which was removed with free margins. The relapse was treated with endocrine treatment, which was changed to tamoxifen, 20 mg daily, and chest wall irradiation was given.

Liquid Biopsy Analysis

The detection and analysis of a panel of circulating rare cell markers derived from 10 mL peripheral blood of the patient is described in supplements (Materials and methods) and in prior work [19, 20, 22, 23]. The marker panel was designed to comprehensively probe MRD dynamic changes in longitudinal testing over a period of 31 months from treatment completion till relapse in a breast cancer survivor and related to the analysis of systemic cancer burden and silent inflammation. The majority of markers have been developed by our group, yet were based on pre-existing knowledge about circulating rare cells [19, 20]. The specific panel of markers and individual function is detailed in Table 1. Further details about cell marker identification are supplemented in Figures S1-4.

Table 1.

Description of the marker panel

Marker	Description	Inflammation character	Pathological Information	References
RPF	Highly active, very immature platelets sized > 5 µm	Pro-inflammation (IL-1α/β, TNF-α, IL-6)	Emergency thrombopoiesis upon thrombopoietin rise, acute platelet loss or cytokine trigger High platelet secretome need (e.g., anticoagulant, anti-ROS) Prediction of pro-inflammatory dominance, if high Prediction of immune-suppression if low (in wound context) Involvement in endothelial dysfunction	[19, 24]
G.Plt	Very large nonnucleated CD31+/DNA- cells sized >12.5 µm	Assumed pro-inflammatory (unknown cytokine milieu)	Emergency thrombopoiesis upon high cytokine levels (assumed trigger mechanism) High platelet secretome need (e.g., anticoagulant, anti-ROS) Prediction of endothelial dysfunction	[19, 25]
CPA1	Small-sized platelet (<3.5 μm) agglutination	Agglutination with change in morphology of pre-activated low sensitivity small platelets	Association with cancer, infection, sterile inflammation, autoimmune, liver and spleen function impairment Platelet exhaustion if high and chronic Sequestration and removal deficiency by spleen/liver Anticoagulant dysregulation (wound tissue)	[19, 26]
CPA2 (clusters)	Larger sized platelet (>3.5 μm) agglutinate	Agglutination of pre-activated, highly sensitive larger immature platelets without change in morphology	Association with cancer and cardiovascular disease Low activation trigger sensitive (e.g., vWF) Low nitric oxide conditions (endothelial dysfunction) Reversible agglutination Agglutination attributed to secretory function Hyperreactivity (e.g., insulin resistance)	So far unreported
nMKc	Viable bare nucleated cell of megakaryocyte after platelet shedding	Size and segmentation increase with increased thrombopoietin and pro-inflammatory signaling (IL-6, Il-3)	Measure of accelerated megakaryopoiesis as consequence of inflammatory upregulation of single cell output or bone marrow hypercellularity	[19, 27]
CEC (including progenitor cells, iCEC)	Spectrum of endothelial cells depending on origin and pathology	Detachment from damaged vessels as a pro-inflammatory factor or bone marrow recruited cells as a repair signal factor	Vascular injury Vascular inflammation Vascular proliferation (angiogenesis, repair) Prediction of tissue hypoxia	[19, 20, 28–30]
CIC	CD44+ rare cell phenotype	Inflammation mediation suspected activation by low molecular weight hyaluronic acid	Representation of altered state of HA turnover (accumulation and depolymerization) Prediction of localized tissue or systemic vascular lesions involving ECM remodeling Measure of oxidative stress	[22, 31]
Normoblast – (nCEB)	Nucleated red blood cell with strong transferrin receptor expression	Pro-inflammatory influenced (Il-6, TNF-α, INF-y, Il-1β) in cancer, or bacterial infections or bone marrow central causes (leukemia etc.)	Impairment of hemoglobin synthesis e.g., iron insufficiency Impaired macrophage retention and increased apoptosis BM erythroid hypercellularity Enucleation failure and block (e.g., hormone imbalance)	[19, 32]
Reticulocyte – enCEB	Large, very immature reticulocyte (CD71 high expression)	Hypoxia/EPO driven and anti-inflammatory regulated	Accelerated stress erythropoiesis due to BM response insufficiency Vascular niche endothelial cell impairment	[19, 33]
CTC	Bone marrow and tumor-derived cells with well-defined epithelial staining	Unclear correlation with inflammation type; CTC egress assumed associated with pro-inflammatory factors	Representation of systemic cancer burden	[19, 20]
EpRC	Non-hematopoietic cell with ill-defined epithelial staining	Unclear correlation with inflammation type	Prediction of all-cause epithelial lesion	[20]
HER2 – HER2-positive cellular event	Non-hematopoietic cell event with positive HER2 staining	No correlation with inflammation status	Measure of systemic cancer burden	Unreported

RPF, rare platelet fraction; G.Plt, giant platelets; RPF, CPA1, circulating platelet agglutinate type 1; CPA2, circulating platelet agglutinate type 2; nMKc, naked megakaryocytes; CEC, circulating endothelial cells; CIC, circulating inflammation-mediated rare cell; CTC, circulating tumor cell; EpRC, epithelial-like rare cells; ECM, extracellular matrix.

Results

Marker Trajectories

Marker trajectories recorded from the studied patient encompassed analysis of platelet disorder. The platelet marker panel included mature platelet count (Plt) as the consequential steady-state, immature large platelets denoted as rare platelet fraction (RPF) and giant platelets (G.Plt) as indicators of stress response to platelet consumption, naked megakaryocytes (nMKc) as indicators of long-term Plt. Steady-state adaptation and circulating platelet agglutinates comprising small (<3.5 μm in diameter) (CPA1) and large platelets (>3.5 μm in diameter) (CPA2) (see Table 1) as reporters of platelet activation. The trajectories are shown in Figure 1.

Fig. 1.

Platelet marker trajectories recorded over the period of post-treatment till relapse lasting 31 months and included markers: very immature large platelets (RPF), giant platelets (G.Plt), naked megakaryocyte (nMKc), circulating small platelet agglutinates (CPA1), and circulating large platelet agglutinates (CPA2).

The diagnostic usefulness of this platelet panel is proposed to relate to the qualification of peripheral low-grade inflammation as a hallmark of tumor microenvironment evolution. Platelet counts in the patient remained within acceptable ranges throughout the observation period, unable to reveal signs of chronic inflammatory stress. A steep and steady increase in emergency thrombopoiesis indicated by RPF and G.Plt was detected until month 19 (RPF: slope 186×, R² = 0.81; G.Plt: slope 9.9×, R² = 0.72), suggesting a dominance in rising systemic pro-inflammatory state and a gradual increase in platelet consumption. In parallel, megakaryopoiesis showed a comparatively slow but steady acceleration until month 19 (slope 0.64, R² = 0.96), consistent with an acute response profile that, however, lasted for several months. At month 19, a transition toward a pro-thrombotic stress profile emerged, inferring stronger platelet activation and consumption: CPA1 exceeded the established normal threshold of 81 events per 5 mL [19] between months 16 and 19, interpreted as wound-activated. By month 21, CPA2 also rose. Therefore, during months 19 and 21, an unresolved inflammatory state was inferred, as neither emergency thrombopoiesis nor bone marrow megakaryopoiesis mounted a compensatory response to the increasing pro-thrombotic stress; instead, both declined and suggested an increase in inflammatory complexity and an underpinning of a tumor microenvironment. After month 21, bone marrow megakaryopoiesis rose with a 4.5-fold stronger upregulation (nMKc slope 2.9×), compared with the earlier trajectory from months 0–19 (slope 0.64×). This behavior was explained in the context of high platelet agglutination and consumption in prior weeks, invoking a rise in thrombopoietin which then increases bone marrow megakaryocyte proliferation.

More so, clinically inapparent low-grade systemic inflammation was further assessed by the impact on bone marrow erythropoiesis. Hereby, circulating erythroblasts (nCEB) and early reticulocytes (enCEB) were used as highly sensitive reporters of stress erythropoiesis, with nCEB representing deficient and enCEB representing hypoxic stress-induced accelerated erythropoiesis [20, 23]. The trajectories are shown in Figure 2. Increased inflammatory tone and hypoxic pressure were suggested when nCEB exceeded 12 cells per 5 mL and enCEB exceeded 19 cells per 5 mL, respectively. Abnormality cutoffs were determined using a healthy cohort as previously published [19].

Fig. 2.

Trajectories of pro-inflammatory markers recorded over the period of post-treatment baseline till relapse lasting 31 months and included normoblast (nCEB) and very immature reticulocyte (enCEB).

In the patient, similar to the findings of platelet stress, stress erythropoiesis seemed to have commenced starting from month 10 (2nd test) and was continued at varying levels until relapse. The inflammatory imbalances seemed to have alternated sequentially between elevation in both nCEB and enCEB, referred to as latent anemia of inflammation (lAI), elevation only in nCEB referred to as latent normoblastemia (lNb) and elevation only in enCEB referred to as latent reticulocytosis. At the 10-months mark, lAI was detected that transitioned to lNb at month 12, inferring erythropoietin (EPO) blunting rather than improved hypoxic conditions, which may be corroborated by the finding of coinciding hypoxia-fueled angiogenesis (iCEC), shown in Figure 3. At month 16, the situation flipped to latent reticulocytosis indicating increased EPO sensitivity and a reduction in erythroblast release, suggesting a decrease in pro-inflammatory impact on the bone marrow. The situation eventually flipped again at month 19 toward lNb-dominated lAI, whereby lAI-positive status continued until relapse, inferring an unresolved inflammatory state. Even though detailed cytokine profiles remain obscure by this method, the platelet and erythroid marker abnormality represents evidence of chronic systemic inflammation with varying impact on the bone marrow occurring already at month 10, suggesting a lead time advantage of at least 20 months when interpreted as a cancer-associated and actionable risk factor.

Fig. 3.

Trajectories of markers representing tissue damage recorded over the period of post-treatment till relapse lasting 31 months and included CD44+ non-hematopoietic rare cells (CIC), inflammation-related epithelial-like cells (EpRC), all type circulating endothelial cells (CEC), and circulating endothelial progenitor cells (iCEC).

Another hallmark of TME evolution is tissue change which, in our panel, was represented by the wound markers including markers of endothelial, epithelial tissue type and extracellular matrix-associated cells (CIC). Vascular health was excluded in the presence of mature and/or abnormal immature circulating endothelial cells (CEC) exceeding a cutoff of a level greater than 2 cells per 5 mL [19]. Angiogenic proliferation was implied when detecting iCEC [20] (for identification, see online supplemented Fig. S2; for all online suppl. material, see https://doi.org/10.1159/000553558). The non-hematopoietic CD44+ rare cells or CIC represented altered matrix protein turnover as a consequence of tissue degradation, transformation, and cell invasion and was described as a novel rare cell phenotype correlating nonspecifically with cancers and diabetes [20, 22]. In more detail, CIC was theorized to mediate inflammatory response to tissue damage based on the immuno-active hyaluronic acid/CD44 interaction similar to CD44+ expressing macrophages, whereby ROS-depolymerized serum hyaluronic acid is assumed to activate the putative CIC [31, 34]. A steady increase in ROS stress was alleged based on the increasing trajectory (R² = 0.78) of CIC over the entire follow-up period. EpCam+ stained non-hematopoietic rare cells with an ill-defined epithelial staining pattern denoted as epithelial-like rare cells (EpRC) were used to interpret findings as epithelial tissue inflammation when exceeding 14 cells per 5 mL (for identification, see online supplementation Fig. S4). The trajectories of the markers to represent tissue damage are shown in Figure 3. A tissue damage profile showing elevation in all three markers was interpreted as a wound. A subcritical wound status was interpreted with elevation in EpRC and additionally in either CEC or CIC. Elevation in one marker was not interpreted as relevant to TME evolution. Upon treatment completion (month 0), the patient did not complain about symptoms at the time and 2–4 weeks thereafter, suggesting absence of acute disease, thus expecting the test result at month 0 to represent the basic physiological status at the given history of adjuvant chemotherapy. At this baseline, rare cell profile abnormality was limited to a slightly positive epithelial wound status (see month 0, Fig. 1–4) which regressed to subcritical status according to the above definition in the second test carried out 10 months later as the component of endothelial injury seemed to have resolved completely (see month 10, Fig. 3). This wound dynamic was nonetheless interpreted as failure to heal without observing significant reduction or even slight increases in EpRC, CIC. The above-noted onset of systemic pro-inflammatory status allows us to suspect an underpinning by chronic wounding (see Fig. 1, 2). A third sample collected 2 months later (at the 12-month mark) revealed a decrease in epithelial inflammation (EpRC) and an elevation of iCEC that we interpreted as wound healing in parts by progressing toward proliferation. However, the continued presence and increase in systemic inflammation (nCEB, RPF, and G.Plt), systemic cancer CTC and oxidative stress (CIC), further supported the previous interpretation of chronicity. A fourth sample taken 16 months after treatment completion substantiated chronicity of the epithelial wound character, having observed a renewed wound-positive status yet with a significant reduction in iCEC. Despite absence of symptoms in the patient, we found it suitable to investigate the clinical tumor status by CT and bone scan for traces of cancer relapse yet returned with negative results. It was suspected that the markers predict only very mild lesions that may lack sufficient density to be detectable by imaging studies. A fifth sample collected at the 19-month mark supported a continued wound profile showing rising epithelial (EpRC, CTC) and endothelial injury (CEC). A sixth sample, collected 3 months later at the 21-month mark, demonstrated signs of healing a second time, indicated by the reduction in epithelial inflammation (EpRC, CTC) and in endothelial damage, leading to the interpretation of subcritical wounding. The 7th and last sample was taken due to the diagnosis of LRR in size of 25 mm and showed absolute peak values in the wound marker panel. At given evidence of tumor growth and when viewed in comprehension, the profile seemed to reveal a fluctuating yet wound-positive status starting at baseline, resembling the common notion of a wound that never heals. However, a clear rising trend of wound status was suggested only for the period between months 21 and 31.

Fig. 4.

Trajectories of markers representing systemic cancer burden or minimal residual disease recorded over the period of post-treatment till relapse lasting 31 months and included epithelial positively stained non-hematopoietic cells (CTC) and HER-2-positive cellular events (HER-2).

For systemic cancer burden analysis, we included all of the previously classified CTC morphologies formerly denoted as CTC1 to CTC6 [20], herein denoted as well-defined CTC. Finally, HER2 signal was included in the analysis and was identified as described in supplemented Figure S3. The trajectories of the markers to represent systemic cancer burden are shown in Figure 4.

An upward trend of cancer-associated cells (epithelial CTC and HER2 events) was obvious in the trajectories of both types, even more with higher consistency by HER2 events achieving a similar slope of 0.43 that translates into a monthly 43% increase in cellular concentration.

MRD in Progression

The ability to detect latent disease progression as a high-risk condition was proposed to be evaluated through the herein presented multi-marker analysis following trajectories toward the recurrence event. Traditionally, progression is understood as a steady increase in tumor mass; accordingly, this progression was commonly correlated with a consistent upward trend in concentrations of cancer-associated markers based on the assumption of continuous tumor growth and invariable shedding and was herein referred to as quantitative progression (QaP). In the patient, only three markers – HER2, nMKc, and CIC complied with the idea of QaP. HER2 displayed a gradual and consistent rise with a slope of 0.27 (R² = 0.82) between months 0–16, and an overall slope of 0.42 (R² = 0.87) over the entire monitoring period, suggesting an average monthly increase of 42%. nMKc showed strong correlation with progression, rising at a slope of 0.65 (R² = 0.96) during months 0–19 and overall at a slope of 1.1 (R² = 0.75) up to relapse, interpreted as a stress response to increasing platelet loss. CIC increased yet in a stepwise fashion: initially averaging 9 cells/5 mL until month 10, rising to 23.5 cells/5 mL between months 12–21, and further increasing to 32 cells/5 mL at tumor onset. Despite its fluctuations, CIC followed an overall positive trajectory with a slope of 0.77 (R² = 0.75). Other markers were less consistent with the model of gradual tumor mass expansion. CTCs showed strong progression between months 10–19 with an exponential slope of 1.2 (R² = 0.99), but only a moderate overall linear trend (slope 0.43, R² = 0.73). RPF and G.Plt showed a similar behavior when compared to CTC, eventually inferring a biological connection. Endothelial clustering as a predictor of tissue hypoxia and angiogenesis was observed only as two transient peaks, suggesting an inflammation-driven, short-term dynamic rather than a marker of steady tumor mass growth. In addition to QaP, we observed a gradual accumulation of the eight TME-relevant pathologies, which we designate as qualitative progression (QlP) (see Supplementary Table S1). These pathologies appeared in a sequential manner and persisted until relapse, with fluctuations reflecting a worsening clinical condition. Notably, a wound-positive status – the key pathology associated with tumor growth – was already evident at baseline. This was followed by the emergence of systemic inflammation and hypoxic stress at month 10. TME-related pathologies involving tissue hypoxia and oxidative stress became detectable at month 12. At the fourth assessment (month 16), the patient tested MRD positive. Platelet suppression was identified at month 21, progressing to excessive platelet loss by month 31. QlP was quantified on a scale from 1 to 8 by assigning values of 0, 0.5, or 1 to each pathology, corresponding to negative, low, and positive status, respectively. At baseline, the patient’s score was 1, while at relapse all pathologies were present, yielding a score of 8. Scores may be categorized into early, advanced, and full progression. Scores 1–3 indicated early progression, potentially reflecting nonspecific pathology including probability of infection; scores of 4–6 denoted advanced progression with evidence of cancer disease in particular CTC; and a score of 7–8 would represent full progression when referring to the evidenced recurrent tumor. Interestingly, the score function, when plotted against the test month, demonstrated high linearity in its progression (R² = 0.97; Supplementary Table S1) and may be consistent with rapid progression in line with the early relapse.

Onset Estimation

The marker trajectories evidence chronicity, predict progression, and may be used to predict onset of tumor growth. The platelet behavior is herein proposed to discern between TME progression that was not detectable by conventional imaging and detectable tumor mass growth. According to the theory, a tumor wound recognizable enough to show signals by imaging based on density contrast such as MRI or PET/CT scans is expected to consume larger amounts of platelets due to tissue irregularity and bleeding. This loss is commonly recognized by lower TPO assimilation and consequently triggers increased megakaryopoiesis. Accordingly, nMKc rise, measured as the rate of increase in cells per month, was identified as one key marker of interest associated with platelet bulk loss. Linear regression analysis of the patient’s nMKc trajectory from month 0–21 revealed a gradual upward trend, with a slope of approximately 0.5 cells per month. This rise was interpreted as a steady increase in megakaryopoietic activity. The slope rose sharply between months 21 and 31 to 2.9 cells per month, inferring a stress response to a substantial rise in peripheral platelet loss. During the initial, gradual growth phase, the lesion was not visible on CT scans, which supports the hypothesis that the TME was evolving until around month 21, as the suggested tumor growth onset. Apart from platelet loss estimation, the 21-month time point was particularly notable, as a slight decrease in nMKc and RPF coincided with a sharp rise in pro-thrombotic markers, suggesting sustained or heightened platelet function demand in the context of diminished platelet supply – potentially reflecting tumor-associated platelet suppression and entailing an altered inflammatory state implying contribution of anti-inflammatory signaling. The RPF/CPA2 ratio effectively captured this behavior quantitatively. At month 21, the ratio dropped sharply, falling from consistently high values – previously exceeding 100 with an average of 534 – to 10, and remained low at 20 upon relapse, suggesting persistence of platelet suppressive signaling.

Discussion

A prospective investigation was carried out in a breast cancer survivor classified as low risk for relapse (T1–2 N0) who underwent mastectomy without post-mastectomy radiotherapy, despite low ER status and relatively large tumor size. The relapse may be a significant setback for the patient, as the 5-year LLR risk is only 6% [3]. The study aimed to assess longitudinal rare cell stability and use the systemic cytology platform to analyze and define characteristics of MRD in the tested patient. Follow-up was intensified in the patient due to high and rising marker trajectories.

Apart from a sedentary lifestyle, conventional risk factor assessment did not predict the patient’s early relapse. While the wound complication theory remains plausible, limited postoperative documentation and the low sensitivity of standard diagnostic methods may have failed to detect subtle inflammation or occult wounds. Such silent inflammatory processes may contribute to, result from, or indicate recurrent disease. The cell-based liquid platform described here enabled the extraction of such diagnostic information and provided evidence that inflammation and signals of tissue change precede the relapse in the patient months prior, suggesting prognostic biomarker value. Overall, signals appeared uncorrelated; however, the evidence of recurrence strongly suggests one underpinning alluding to TME evolution as such inferring a more complex yet strong pathological association and interdependence between all marker trajectories.

The dormancy/reactivation model to explain recurrence was supported by our data. The trajectory of CTCs as the key marker for MRD detection was too dynamic over the follow-up period to be interpreted as latent, indolent, or dormant. Systemic inflammation seemed to have followed a cyclic pattern based on the trajectory of hypoxic conditions and may support the theory of subclinical infections periodically flaring up as proposed relapse causation [12, 13]. The enCEB marker served as a primary marker to evidence inflammation cycles based on observed falling and rising trends. Even though the interplay between systemic hypoxic stress and bone marrow inflammatory signaling is complex and a bone marrow biopsy was not measured, speculations about the marker trends may support the interpretation of cyclical chronic inflammation. The enCEB elevation in the setting of lowered nCEB levels (month 16) could be explained by mounting hypoxic stress producing peak EPO signaling that in consequence enhances erythroid cell survival and promotes progression toward terminal differentiation, whereas residual inflammatory influences may limit the number of erythroblast proliferative cycles, thereby causing dominance in early reticulocyte formation. In analogy, high nCEB and low enCEB (month 21) were explained by a lowering in systemic hypoxic stress and EPO signaling and increased bone marrow centric pro-inflammatory signaling, eventually leading to lowered bone marrow retention and increased normoblast apoptosis [19, 35]. The “infection theory” may also be supported by the rise in iCEC observed at month 12, suggesting increased endothelial proliferation and localized angiogenic and hypoxic processes followed by resolution, a pattern that may resemble the vascular response observed during acute infection. The notion of repeated shifts in inflammatory signaling that failed to resolve may be consistent with a driving force behind TME evolution – acting as continuous or sequential triggers of cancer evolution, rather than a single reactivation event.

The analysis demonstrated the feasibility of the platform to detect MRD positivity based on CTC elevation when above cutoff [20] with a lead time advantage of 15 months. MRD positivity based on liquid biopsy commonly achieves a significant lead time advantage, paving the way for interventions eventually years before clinical relapse, providing more precise prognostic insights [8, 36], and may also function as a surrogate endpoint in clinical trials. The prognostic value of CTC and ctDNA is commonly accepted [8, 37–39]. Although sensitivity may remain modest for CTC and ctDNA, the likelihood of relapse, once detected, has been reported to be certain [10, 40]. A key limitation of early MRD detection is the uncertainty associated with marker persistence, optimal treatment selection, and the timing of clinical relapse. This uncertainty may increase patient anxiety, adversely affect quality of life and mental health, and complicate treatment planning [41]. This unpredictable time frame leads to a “wait-and-see” state that necessitates ongoing, potentially long-term monitoring, placing additional burdens on healthcare resources and increasing costs for both patients and health systems. The situation is aggravated by fixed or inadequately data-informed follow-up scheduling commonly ranging from 1 to 6 months, dealing inadequately between oversampling and missing the tumor onset. Naturally, markers are anticipated that enable prediction of MRD in progression and estimation of time-to-relapse with reasonable accuracy, as well as allow optimized test interval determination. A solution denotes the determination of MRD progression with the inherent information of trajectories eventually offering crucial timing insights, enabling differentiation between stably low, indolent, or actively progressing MRD and consequently guiding the prioritization of interventions and monitoring intensity. While predicting MRD progression using ctDNA has proven feasible in leukemias to guide clinical management [42], applying CTCs – particularly in solid tumors – remains challenging. Although CTCs expressing EpCam and cytokeratin are technically reliable due to consistent detection and sensitivity to low-level cancer signals, their biological reliability for disease monitoring has not been established [8]. EpCAM can be downregulated during EMT, meaning that disease progression could actually occur without EpCAM signals. More so, EpCam/ cytokeratin CTC showed a low specificity toward cancer risking false interpretation [43–47]. Our findings of CTC trajectory are consistent with reported fluctuations or even positive-negative flips in prior studies, leading to unacceptable uncertainty in progression trend analysis [40]. We may propose the herein presented multi-marker panel analysis as a potential approach to increase marker trajectory reliability, leveraging both qualitative (pathology interpretation-based) and quantitative (single marker absolute value) information of progression. The herein proposed marker trilogy – epithelial and endothelial tissue injury along with matrix protein turnover changes (EpRC, CEC, and CIC) denoting a profile pathology for wound prediction was found reliable but not consistent with the idea of steady mass increase. Both EpRC and CEC greatly fluctuated over the measurement period, failing to show a clear rising trend and suggesting a closer link to inflammatory patterns. This may also be supported by the biological context of EpCam, which is inflammation specific and not cancer specific [48]. Only three markers showed consistency with the idea of progressing wound growth, comprising HER2, nMKc, and CIC. While nMKc and CIC could be explained by their biological function mentioned earlier in the results section, HER2 progression was unexpected since both the primary and recurrent tumor were found HER2 negative by conventional laboratory tests. The finding may represent a false-positive identification attributable to autofluorescence in the red emission spectrum with a strong correlative pattern with cancer evolution. Alternatively, it could reflect the well-documented phenomenon of HER2 discordance in liquid biopsy, reported in up to 50% of cases and frequently linked to poorer prognosis [49–51]. Admittedly, the marker may not be useful in the majority of cancer patients, denoting a relatively rare finding even in breast cancer patients. Accordingly, nMKc and CIC are proposed here as key markers of QaP. Conceptually, follow-up intervals within the commonly used range of 1–6 months could be individualized according to combined QaP and QlP progression scores. In that sense, lower progression scores may support longer monitoring intervals, whereas higher progression scores could justify intensified surveillance. With larger datasets and prospective validation, the MRD progression markers may potentially contribute to estimating time-to-relapse. In application of the current data set, an increase of 0.77 cells per month in CIC and a 2.6-cell increase in bone marrow megakaryopoiesis per month and an increase of 0.22 abnormalities/month would predict a maximal 15 months lead time.

Moreover, our data set suggests that shifts in the inflammatory wound profile and evidence of platelet loss may also be informative of tumor mass growth onset. Pro-inflammatory signaling is thought to suppress relapse by maintaining dormancy and preventing apoptosis, whereas anti-inflammatory dominance may favor tumor regrowth. Consistent with this, we observed an inflammatory signal shift between months 19 and 21, marked by a sharp decline in the RPF/CPA2 ratio (from ∼500 to ∼20) largely due to a decline in RPF. This anti-inflammatory shift coincided with a rise in MKc levels (0.4× to 2.9×), indicating platelet consumption believed to be caused by a tumor wound site. Together, these events were considered to signal recurrence onset. Of note, the CTC drop at the 21-month mark and the inflammation-driven RPF trajectories appeared similar, suggesting dependency of CTC egress on inflammatory signaling.

Conclusion

A multi-rare cell marker liquid biopsy platform, applied to a borderline low-risk breast cancer survivor, corroborated previous reports on the feasibility of cell-based MRD detection, demonstrating a lead time advantage of several months before relapse. In addition to identifying MRD positivity, the results demonstrated longitudinal evidence of MRD progression. Accordingly, the multi-marker approach may serve as a clinically informative marker of MRD progression, thereby enhancing diagnostic confidence and supporting clinical decision-making and the effectiveness of intervention. In particular, HER2 events, nMKc, and CD44⁺ rare cells exhibited longitudinal patterns consistent with increasing disease severity and tumor progression. The case results may also improve understanding of tumor-host interactions and recurrence biology. Our findings do not align with a dormancy or latency/reactivation model; instead, they suggest an active, clinically inapparent, yet gradually progressing inflammatory disease state.

Statement of Ethics

This study was performed in accordance with the Declaration of Helsinki. This human study was approved by the Mahidol University Central IRB, Mahidol University – Approval: 2019/197.3007. All adult participants provided written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for publication of the details of their medical case and any accompanying images. The CARE Checklist for case reports has been completed and is included as online supplementary material.

Conflict of Interest Statement

The corresponding author and Suparerk Borwornpinyo are shareholders of a company involved in the development and manufacturing of cell separation technology and biomarkers during the conduct of this work. The remaining authors have no conflicts of interest to declare.

Funding Sources

The authors declare that they received no funds for this research.

Author Contributions

S.S. conceived and designed the study, carried out imaging analysis, and drafted the manuscript. P.rB. carried out experimentation and took part in image analysis. S.B., P.C., and W.T. advised and reviewed the study. P.L. reviewed the study and took part in drafting the manuscript. All authors read and approved the final manuscript.

Funding Statement

The authors declare that they received no funds for this research.

Data Availability Statement

The datasets generated and/or analyzed during the current study are not publicly available due to reasons of sensitivity but are available from the corresponding author on reasonable request.