Breast cancer remains one of the most challenging and predominant malignancies worldwide, requiring innovative ways to understand its progression and improve early detection schemes (1, 2). It initiates with genetic mutations, which lead to uncontrolled cell growth, eventually forming a tumor inside the breast. This evolution involves various stages, which include hyperplasia, carcinoma in situ, invasive carcinoma, and metastasis (3, 4). Recent studies have highlighted the role of cancer stem cells in tumor origination and development (5–7).

Early detection methods, such as mammography, biomarker analysis, and MRI, are critical for detecting breast cancer at its most treatable stage. Additionally, a large-scale study comparing the effectiveness of digital and film mammography in early breast cancer detection found that digital mammography is particularly beneficial for younger women and for those with dense breast tissue (8). An individual may notice signs or symptoms of breast cancer under standard care, prompting them to seek for medical evaluation and eventually receiving a proper diagnosis. In contrast, earlier detection occurs when an individual without any noticeable signs or symptoms is diagnosed through a specialized screening examination (9). Recent evolving technologies, such as AI-assisted imaging and liquid biopsies, show potential in improving the accuracy of early detection (10).

Lee and Zelen (2003) studied a stochastic model for predicting breast cancer mortality, focusing on the effects of early detection and treatment. The model estimates survival probabilities based on tumor progression and the timing of diagnosis, providing insights into the impact of screening programs on breast cancer outcomes. It further found that early detection of breast cancer through screening significantly reduces mortality by identifying tumors at a more treatable stage (9, 10).

The immune system plays a crucial role in recognizing and eliminating cancer cells (11). Schreiber et al. (12) reviewed the mechanisms of cancer immunoediting, demonstrating how the immune system can both suppress and promote tumor growth, and highlighting the importance of a balanced immune response, where key players include cytotoxic T lymphocytes (CTLs), natural killer (NK) cells, and macrophages. Recent research has also focused on the role of regulatory T cells and myeloid-derived suppressor cells in modulating the antitumor immune response (13, 14).

Mathematical modeling has emerged as a powerful tool in oncology, offering insights into tumor growth dynamics and the complex interactions between cancer cells and the immune system (11, 14–22). The application of mathematical models to cancer research has a rich history, dating back to the seminal work of Norton and Simon in 1977, who proposed a Gompertzian model of tumor growth (23). Since then, numerous models have been developed to capture various aspects of cancer biology, including the pivotal role of the immune system in tumor suppression and progression (23, 24). Recent advancements in early detection technologies have significantly improved breast cancer outcomes. However, the precise impact of early detection on tumor–immune dynamics remains incompletely understood.

Mathematical modeling offers a unique opportunity to explore these interactions and potentially optimize screening protocols (25). These models range from simple ordinary differential equations to complex agent-based models and have been instrumental in predicting treatment outcomes and optimizing therapy schedules (26). The study in (26) emphasized the importance of collaboration between mathematicians, biologists, and clinicians in developing and validating predictive models for cancer treatment. While optimal control theory provides a framework for personalized treatment, more research is needed to develop and validate personalized models that account for individual patient variability.

Optimal control theory involves determining the best intervention strategies to achieve desired outcomes, such as minimizing tumor size while maximizing immune response. Lenhart (27) applied optimal control theory to biological models, illustrating how this approach can be used to design optimal treatment schedules for cancer patients, balancing efficacy and side effects. This approach has been applied to various cancer treatment modalities, including chemotherapy, immunotherapy, and radiotherapy (28). The study in (29) explores the interactions between tumor cells and the immune system, including immune surveillance, immune evasion, and immunoediting.

Recent advances in this field have led to the development of the cancer immunogram, a framework for understanding the multifaceted nature of tumor–immune interactions (30). Stability analysis involves examining the equilibrium points of the ODE system to understand the conditions under which the system remains stable or undergoes changes, such as tumor regression or progression (31–34). Advanced techniques, such as bifurcation analysis, have been applied to study the qualitative changes in system behavior under varying parameters (34–36).

More recently (37), analyzed long-term trends in breast cancer incidence, concluding that improved screening has led to increased detection of early-stage cancers but has had a limited impact on the incidence of advanced disease. Building on the work of Schreiber et al. (38), conducted a comprehensive analysis of the tumor microenvironment, revealing the prognostic significance of various immune cell subsets in different cancer types.

Recently (39), discussed the integration of mathematical modeling with clinical data to develop personalized treatment approaches in oncology. Building on this foundation, Sharifi et al. developed an optimal control framework for combination immunotherapy, demonstrating how mathematical modeling can guide the design of more effective treatment protocols (40).

Combining early detection methods with an understanding of the immune response provides a comprehensive strategy for managing breast cancer. Early detection is critical for identifying treatable stages of the disease, while the immune response plays a significant role in controlling tumor growth. Recent studies, such as those by (41), have highlighted the potential of using immune biomarkers in conjunction with traditional screening methods to improve early detection and prognosis prediction.

Modeling and optimal control theory offer a structured approach to determining the best intervention strategies, balancing efficacy and safety. This approach can lead to personalized treatment plans that maximize patient outcomes (42–44). Recent work by (45) has demonstrated the potential of using optimal control theory to design adaptive therapy protocols that may delay or prevent the onset of drug resistance in cancer treatment. Recent advances in single-cell sequencing technologies offer new opportunities to characterize tumor heterogeneity and develop more precise models (41, 45).

Advancing the field requires collaboration between oncologists, immunologists, mathematicians, and data scientists. Such interdisciplinary efforts can lead to the development of more sophisticated models and effective treatment strategies. Initiatives like the Physical Sciences Oncology Network (PSON) are fostering such collaborations, but more work is needed to bridge the gap between theoretical models and clinical practice (46). While many mathematical models show promise in simulating cancer dynamics, there is a need for rigorous validation using clinical data.

The work of (39) provides a framework for integrating mathematical modeling into clinical trials, but wider adoption of these approaches is needed. As our understanding of cancer biology evolves, mathematical models need to be updated to incorporate new insights. For example, recent discoveries about the role of the microbiome in cancer progression and treatment response by (46) present new challenges and opportunities for model development. As models become more complex, incorporating multiple scales and high-dimensional data, there is a need for advanced computational methods to solve and analyze these models efficiently (42).

Mathematical modeling has played a vital role in understanding the progression and treatment of breast cancer. Several studies have proposed models capturing key aspects of tumor dynamics, treatment response, and microenvironmental interactions (39, 42).

Benzekry et al. developed a mathematical model using systems of ordinary differential equations to explore and investigate systemic interactions between primary and metastatic tumors, focusing particularly on breast cancer. Their work demonstrated that distant tumors could exert inhibitory or stimulatory effects on each other’s growth. This suggests the need for a systemic view of tumor control beyond localized therapy (47). The role of cancer stem cells in breast cancer response to radiotherapy, using a spatial mathematical model, was examined in (48), highlighting the significance of the spatial distribution of stem-like and proliferative cells within a tumor. Their findings showed that treatment outcomes were not only dependent on the quantity of cancer cells but also on their spatial organization and phenotypic heterogeneity.

Pérez-García et al. (49) proposed a nonlinear mathematical model to analyze the combined effects of chemotherapy and radiotherapy on tumor growth. Although the model is not limited to breast cancer, it is generalizable and applicable to solid tumors. The study generally offered insights into the design of optimal treatment schedules and conditions for tumor eradication, underscoring the impact of treatment timing and intensity on tumor dynamics.

An agent-based modeling (ABM) approach is now being employed to simulate the progression of ductal carcinoma in situ (DCIS), a common form of early-stage breast cancer (50). Their multi-scale model integrated cellular behaviors with tissue-level structures, enabling the study of spatial tumor growth, intraductal invasion, and treatment effects. This work emphasized the utility of in silico platforms in understanding the heterogeneity and morphology of early breast lesions. Macklin and Lowengrub focused on the influence of the tumor microenvironment on breast cancer development. They introduced a hybrid model that coupled mechanical pressure, nutrient diffusion, and extracellular matrix interactions. Their simulations revealed that tumor shape and invasion patterns were significantly affected by microenvironmental conditions, demonstrating the importance of spatial and physical factors in tumor progression (51).

More recent efforts have extended these frameworks using advanced methodologies. Dołęga-Kozierowski et al. integrated numerical modeling with AI and image fusion to produce personalized 3D breast cancer models for enhanced prediction of tumor growth and spatial behavior (52). A virtual twin computational model that predicts neoadjuvant therapy outcomes in individual patients based on imaging and clinical data was introduced in (53). These studies illustrate a growing trend toward data-driven, patient-specific modeling. Li and Thirumalai presented a model focusing on phenotypic heterogeneity, particularly HER2 status, and showed how this influences treatment sequencing and resistance (54). In a related study, Yang et al. developed a model to track breast cancer cell population dynamics under doxorubicin exposure, quantifying drug resistance at the cellular level (55). Idrees et al. applied fractal–fractional calculus to breast cancer dynamics, enabling more accurate depiction of memory and nonlocal effects in tumor growth (15).

Recent advances in early detection and tumor–immune interaction modeling have highlighted the importance of integrating biological data and modern computational tools to improve breast cancer diagnosis and treatment outcomes. Zeng et al. (56) demonstrated the potential of combining serum Raman spectroscopy with convolutional neural networks for rapid and accurate identification of breast cancer subtypes, providing a powerful tool for early detection. Pang et al. (57) investigated the prognostic significance of systemic immune-inflammation indices in HER2-positive metastatic breast cancer, emphasizing the critical role of immune markers in clinical decision-making. Similarly, Liu, Zhang, and Zhao (58) identified transmembrane protein 100 as a biomarker for malignant progression and chemosensitivity, underscoring the molecular basis for individualized treatment strategies.

From an immunotherapy perspective, Zhu et al. analyzed the efficacy of PD-1 inhibitor therapy combined with the GP regimen in advanced triple-negative breast cancer, reflecting current trends in combining immune checkpoint blockade with chemotherapy (59). In addition, Cui et al. (60) developed a nomogram to predict radiation-induced dermatitis, aiding in personalized radiation treatment planning. Advances in artificial intelligence have also contributed to model-informed oncology, where it is applied as causal representation learning for radiology report generation, bridging imaging and clinical semantics in cancer care (61).

Despite the richness of existing models, most do not explicitly incorporate the role of early detection and public awareness in controlling tumor growth. This is a crucial gap given the known benefits of early diagnosis in reducing breast cancer mortality. The model proposed in the present study addresses this by introducing an awareness-driven early detection mechanism, mathematically embedded into the tumor growth dynamics. This novel approach allows for quantitative evaluation of how awareness campaigns, screening programs, and public health interventions can influence cancer outcomes—an area not sufficiently explored in previous literature.

We adopt the mathematical model in (17) and incorporate an awareness term that signifies early detection of breast cancer to stress the importance and effect of early detection in the treatment of breast cancer. The motivation for this research is driven by the urgent need to enhance early detection methods and treatment strategies for breast cancer. By bridging the gap between theoretical modeling and clinical practice, this study aims to provide valuable tools and insights that can inform both researchers and healthcare professionals in their efforts to reduce breast cancer mortality and improve the quality of life for patients.

This study is organized as follows: Section 1 provides the introduction. The model is presented in Section 2, while Section 3 deals with the stability and sensitivity analysis of the model. We introduce the optimal control problem and its analysis in Section 4. We present the numerical simulations in the last section, followed by the conclusion and discussion.

The model in (17) is given in (Equation 1) as follows:

Where C ( t ) represents the population of breast cancer cells at time t , L ( t ) represents the population of cytotoxic T lymphocytes, a 1 is the tumor intrinsic growth rate, a 2 is the carrying capacity for breast cancer cells, a 3 represents the rate at which tumor cells are killed by lymphocytes, a 4 is the saturation constant for the killing of tumor cells by lymphocytes, a 5 is proliferation rate of lymphocytes, a 6   is carrying capacity for lymphocytes, a 7 represents saturation constant for lymphocyte proliferation, a 8   is the rate of lymphocyte death due to cancer cells, and a 9 is the natural death rate of lymphocytes due to apoptosis. Our mathematical model given in (Equation 2) is an extension of (17) by incorporating the early detection term as follows:

Where f ( C ) is the effect of early detection, which is a function of C ,   the tumor size. This effect is influenced by tumor size C ( t ) , detection threshold C ˜ , and awareness parameter α . The function f ( C ) expressed in (Equation 3) is considered to follow the Michaelis–Menten term (62–66). Thus, it is given as follows:

where K represents the maximum effect of early detection, C ( t ) is the tumor size at time t ,   C ˜ represents the tumor size threshold for detection, a is a constant that ensures the function behaves properly for small C ( t ) by providing a baseline detection effectiveness, and α is the awareness parameter. We can observe that high awareness ( α ) makes early detection more effective, lowering the overall impact of tumor size. Conversely, low awareness makes early detection less effective, allowing the tumor to grow more before detection has a significant effect. This term ensures that as C ( t ) grows, detection saturates, reflecting that detection effectiveness reaches a maximum point.

For ease of analysis, we nondimensionalize our model to reduce the number of parameters involved. The nondimensionalized model is given by:

where,

In this section, we conduct a sensitivity analysis, an important technique for assessing how the model’s solution responds to changes in its parameters. We examine the sensitivity of tumor concentration to each parameter, evaluating it through forward sensitivity index analysis.

Figure 1 presents the graph of the sensitivity analysis. The bar plot illustrates the sensitivity of tumor concentration x ( T ) to each parameter. Parameters shown in green reduce tumor concentration when increased, whereas those in red cause it to rise. This visual quickly allows identification of the parameters are most influential for control strategies. The most influential parameter in reducing tumor is κ (effect of early detection), followed by α (awareness) and ( β ,   ρ ) immune response. Enhancing awareness and immune response reduces tumor concentration, as shown in Figure 1 . High immune decay ( η ) or low immune effectiveness ( γ ) increases tumor burden. Table 1 provides a detailed description of the sensitivity analysis.

In this section, we use the parameter values from Table 2 to present the numerical simulations of model (2) showing the impact of early detection on tumor growth, using MATLAB version 2017b to draw Figures 1 – 12 (25, 78, 79).

Figure 2 presents a time-series comparison of tumor concentration with and without early detection. The tumor shrinks rapidly to nearly zero, indicating that early detection is highly effective in controlling tumor growth. However, without early detection, tumor concentration stabilizes at a higher level, indicating persistent tumor presence. This graph clearly illustrates the significant role of early detection in reducing tumor burden.

Figure 3 shows the value of the early detection term, which is a function of tumor concentration at various levels. This graph does not show the tumor level but rather the effectiveness of the early detection term. The blue curve ( α   =   0.05 ) lies higher, meaning that at every level of x , the early detection term is stronger and removes more tumor. In contrast, the purple curve (α = 0.8) lies lower, meaning that at every level of x , the early detection term is weaker and removes less tumor. When small α (i.e., high awareness) implies a stronger early detection term, which leads to greater tumor reduction. Conversely, a large α (i.e., low awareness) implies a weaker detection term, resulting in less effective tumor reduction. In general, stronger early detection leads to greater tumor reduction.

Figure 4 present a bifurcation diagram illustrating how the equilibrium values of the system’s variables change as the parameter β (tumor–immune interaction rate) varies. The blue curve fluctuates dramatically, suggesting complex tumor dynamics as β increases. Multiple bifurcation points may occur, where small changes in β lead to large shifts in tumor’s steady state. Tumor concentration values are on the order of 10 − 6 , indicating very small steady-state tumor populations; however, they are not identically zero, showing persistence of tumor even under high immune killing.

The green curve, representing immune cells, is almost flat and near zero, indicating minimal change in immune cell equilibrium with varying β . This might suggest immune suppression or exhaustion in the model. Sharp oscillations in the curve indicate nonlinear, and possibly chaotic, tumor responses to changes in β . The bifurcation analysis used the cancer-free equilibrium E 0 = ( 0 ,   0 ) , and the observed transition illustrates a forward (supercritical) bifurcation with respect to β .

Bifurcation analysis with respect to the awareness level α is shown in Figure 5 . As the awareness parameter increases, the tumor equilibrium decreases. This indicates that a higher level of awareness, more effective campaigns, or earlier screening significantly reduce tumor size. There is no sharp bifurcation point; however, there is a continuous decline in tumor equilibrium.

Figure 6 referred to bifurcation analysis with respect to the early detection strength κ . As κ increases, the tumor equilibrium value decreases. This implies that stronger early detection significantly suppresses tumor growth. The transition appears smooth, without any abrupt bifurcation point.

Figure 7 is a 3D graph of time (x-axis), awareness level (y-axis), and tumor size (z-axis). At low awareness levels (left side of the y-axis), the tumor starts at a certain size and grows rapidly over time (higher z-values). When awareness increases (moving right on the y-axis), the tumor shrinks more quickly. This implies that at high awareness levels, the tumor is detected early and controlled, leading to the tumor size dropping to almost zero early and remaining suppressed throughout the time span.

Figure 8 shows the phase portrait of Tumor–Immune Dynamics, where the x- and y-axes represent the concentrations of tumor cells and immune cells actively engaged in tumor suppression, respectively. The trajectory illustrates how the state of the system evolves over time in the phase space defined by tumor size and immune response. It starts with a low tumor burden and low immune activity, and as time progresses, the system moves along the path shown.

Biologically, this indicates that both tumor cells and immune response start at low levels, representing a situation immediately after tumor initiation, when the tumor is small and not yet strongly detected by the immune system. The progression along the curve implies that as tumor size increases slightly, the immune system begins to activate. The upward trajectory suggests that the immune system is responding proportionally to the tumor presence. The curve indicates a positive, nonlinear relationship between tumor size and immune response. This suggests that the immune system becomes increasingly responsive as tumor cells proliferate. This aligns with known biological phenomena, in which the immune system often remains inactive until the tumor burden exceeds a certain threshold that triggers recognition and response. The curve gradually flattens, indicating that even as tumor size grows, the immune response increases more slowly. This indicates immune exhaustion or checkpoint inhibition effects. It also represents suboptimal immune activation, where increasing tumor size is not met with a proportionate immune escalation.

The basin of attraction in Figure 9 shows how the final tumor concentration depends on the initial conditions of the system—specifically the initial tumor cell population x ( 0 ) and initial immune cell population ( 0 ) . When y ( 0 ) is moderately high (e.g., >0.5) and x ( 0 ) is not too high, the tumor is often reduced to low final levels (dark purple). This indicates that immune cells successfully suppress tumor growth. Light yellow zones indicate initial conditions where the tumor persists or grows, possibly due to low initial immune presence, a large initial tumor load overwhelming the immune system, or insufficient early detection ( κ ,   α ) .

Figure 9 underscores the importance of early immune intervention and tumor detection. There are immune presence thresholds below which the tumor cannot be controlled. This supports the value of early detection campaigns and strategies to boost immune function in breast cancer treatment.

Figure 10 gives the phase portrait with vector fields, nullclines, and trajectories for system (3). The vector field (gray arrows) shows the direction of movement or dynamics of the system (small arrows). It illustrates how the state of the system changes over time. For example, at a point ( x , y ) , the arrow indicates how tumor and immune levels change next.

For the nullclines, we can observe the tumor nullcline, where tumor size does not change (blue line), the immune nullcline (red dashed line), and their intersection (black dot), which gives the coexistence steady state. The coexistence steady state, as shown in the graph, is (0.25, 0.70). If the system reaches this point, it will remain there. This tells us that the tumor stabilizes at 25% maximum size, while the immune response stabilizes at 70%. Each line represents a simulation from different initial conditions (IC) described in Table 3 .

Observe that all trajectories converge toward the equilibrium point, confirming local stability. The direction of flow confirms that the system naturally regulates tumor growth, depending on the strength of the immune response. In conclusion, this phase portrait demonstrates stable tumor–immune dynamics: the immune system can contain the tumor at moderate levels, depending on initial conditions, and the system is locally stable around the equilibrium point (0.25, 0.7), validating our stability analysis.

Figure 11 compares the evolution of immune cell concentrations over time under two scenarios—with and without early detection. Early detection reduces tumor burden, which in turn lowers immune stimulation and causes a decrease in immune cell levels. In contrast, the absence of early detection results in a more active tumor presence, which sustains higher immune response levels over time. This figure clearly demonstrates the impact of early detection in reducing immune activation demand, thereby supporting the effectiveness of early cancer diagnosis in controlling immune–tumor dynamics.

Figure 12 demonstrates the model fitting of tumor concentration over time. The blue line represents the mathematical model prediction x ( τ ) , and the red dots represent synthetic real tumor data. The model provides a good fit, effectively capturing tumor growth dynamics over the observed time period.

Due to the lack of access to patient-specific tumor concentration data, the data used for model fitting are synthetic, designed to emulate realistic growth trends. This approach is supported in the literature as a valid method for validating mathematical oncology models (16, 25).

We consider a mathematical model of breast cancer with cytotoxic lymphocytes, incorporating the effect of early detection in treating the disease. The model was nondimensionalized to simplify analysis by reducing the number of parameters involved. Two steady states were obtained: cancer-free state and a coexistence states. Local and global stability analyses were conducted using a suitable Lyapunov function, and conditions for stability was stated. A sensitivity analysis was conducted to determine how sensitive the solution of the model is to changes in parameters.

We explored the sensitivity of tumor concentration to each parameter using the forward sensitivity index analysis. Figure 1 shows the sensitivity analysis graph, where the most influential parameter in reducing tumor is κ (effect of early detection), followed by α (awareness), and β ,   ρ (immune response). Improving awareness and immune response helps reduce tumor concentration, as observed in Figure 1 , while high immune decay ( η ) or low immune effectiveness ( γ ) worsens tumor burden. A comparison of tumor concentration with and without early detection shows that the tumor shrinks rapidly to nearly zero, indicating that the early detection mechanism is highly effective at controlling tumor growth. However, tumor concentration stabilizes at a higher level, indicating its persistent presence due to the lack of early detection intervention. These results clearly demonstrate the significant role of early detection in reducing tumor burden.

More so, bifurcation diagrams were presented with regards to β (immune response), κ (early detection parameter), and α (awareness campaign). As the awareness and early detection parameters increase, the tumor equilibrium decreases, indicating that higher levels of awareness, more effective campaigns, or earlier screening significantly reduce tumor size. Figure 9 shows the basin of attraction, emphasizing the importance of early immune intervention and tumor detection. There are thresholds of immune presence below which the tumor cannot be controlled. This supports the value of early detection campaigns and immune function enhancement as clinical strategies in breast cancer treatment.

In conclusion, despite the threat, severity, and deadly nature of breast cancer, early detection and awareness campaigns are key to curing or improving patients’ lives, as indicated by the mathematical model and supported by numerical simulations in this study. Thus, relevant authorities should emphasize the need to strengthen early detection and awareness campaigns among the populace to curb the menace posed by this deadly disease.

We believe that the current model provides valuable insights into the interplay between tumor dynamics and immune response, particularly under the influence of awareness-driven early detection; however, it has some limitations:

To address these limitations and enhance the model, future research could explore the following avenues: