The incidence of tricuspid regurgitation (TR) increases with age, and even less severe cases are associated with poor prognosis in terms of morbidity and mortality.1 Until recently, the only available treatment option was surgery, which was, nevertheless, rarely performed due to the high mortality rate among patients, who are typically in advanced stages of the disease and present with multiple comorbidities. Advances in percutaneous therapies have provided a viable alternative to surgery for selected patients with severe symptomatic TR and have considerably expanded our understanding of the disease.

In this context, an anatomical classification of tricuspid valve types has been developed, along with a classification system for TR based on its underlying mechanisms, taking into account the prognostic impact of the disease. The growing number of patients treated with various percutaneous devices has further contributed to this knowledge, demonstrating that these therapies are safe and associated with improved quality of life. However, since these patients are often elderly and have undergone prior interventions or present with significant comorbidities, risk stratification remains challenging. Moreover, the optimal timing and type of intervention required to modify disease progression and improve survival are still uncertain, and robust clinical evidence to guide these decisions is lacking.

To better classify these patients, numerous studies have explored prognostic stratification based on TR type, imaging parameters of right heart remodeling,2–4 clinical indicators,5 and laboratory markers of congestion.

The TRI-SCORE risk scale, designed to estimate mortality rates associated with isolated tricuspid valve surgery, is the most extensively studied to date.6 After external validation, TRI-SCORE was applied in a French registry of patients with severe TR undergoing surgery, percutaneous treatment, or medical therapy. It proved effective in stratifying patients into low-, intermediate-, and high-risk groups and in assessing the potential benefit of treatment based on that risk.7 Although simple and easy to use, the scale does not include newer clinical or advanced imaging variables that have demonstrated prognostic value in recent studies. Furthermore, its utility in the clinical follow-up of patients with moderate or severe TR has not been evaluated.

In recent years, several studies have shifted focus from estimating procedural risk to identifying factors that influence disease progression and prognosis. With support from artificial intelligence (AI), these studies have attempted to classify patients into distinct groups based on their clinical characteristics, TR type, and findings from physical examination, laboratory testing, and imaging.8–10 This type of technology may be the only feasible approach to comprehensively capturing the range of disease-related prognostic factors and identifying distinct risk profiles. In this context, the study by Badano et al.,11 recently published in Revista Española de Cardiología, used unsupervised cluster analysis to identify prognostic phenogroups of secondary TR (STR), integrating advanced imaging techniques and machine learning. This approach assessed right atrial and ventricular size and function using 3-dimensional imaging and speckle-tracking technology. As a result, 3 risk phenogroups—low, intermediate, and high—were identified with high accuracy and reliability in both derivation and validation cohorts. This study reinforces the concept of STR as a heterogeneous condition and offers a quantitative framework for risk stratification with direct clinical implications.

First, the study highlights the value of advanced echocardiography (3D imaging and speckle-tracking analysis of ventricular mechanics) as both a differentiating tool and a determinant of prognosis. It emphasizes the importance of right heart remodeling as a critical factor in managing these patients. Notably, patients with only moderate TR had a high incidence of the composite endpoint and death, indicating the need for risk stratification that goes beyond simple severity assessment. Advanced echocardiography is particularly valuable in intermediate-risk patients who might otherwise be classified as low-risk using standard imaging techniques. Early detection of right-sided dysfunction and remodeling allows for timely reclassification into a higher risk category.

Second, the study provides robust prognostic data in a well-characterized cohort of patients with significant TR, all under rigorous follow-up. It confirms a high event rate over a short period, even among low-risk patients with non-severe TR, reinforcing the necessity of including right heart remodeling in risk assessment. Given these high event rates, it is essential to evaluate the impact of available treatments on morbidity and mortality. We must determine the extent to which these treatments alter the disease course and identify which patients are most likely to benefit—thus avoiding futile interventions.

Third, the study proposes a model that is applicable to clinical practice, having been validated in an external cohort. The strong performance in the validation group supports the model's robustness and its potential broader use for risk stratification in patients with significant TR.

Lastly, it is important to highlight the role of artificial intelligence in achieving a comprehensive view of all factors influencing disease prognosis. Traditional scoring systems are necessarily simplified to enable their use in clinical care. We are often required to select certain variables, and in most cases, to interpret them dichotomously for ease of integration, which leads to a loss of valuable information about our patients. Disease exists on a continuum in which numerous clinical, laboratory, and imaging parameters interact to define each patient's phenotype and prognosis. Artificial intelligence helps us retain this complex information, making it an essential element for more precise, individualized management aimed at achieving the best possible health outcomes.

Therefore, the findings of this study could transform STR management by improving risk stratification and optimizing patient selection for therapeutic interventions. In addition, they could inform the design of future clinical trials and support the integration of AI into routine clinical practice.

Despite the strengths and promising findings of the study, it has some limitations, such as its retrospective design and the complexity of the evaluation process. Future research should aim to prospectively validate the proposed phenogroups, assess the impact of specific interventions within each group, and develop simplified tools for patient classification. Furthermore, the model lacks data on biomarkers and systemic comorbidities (eg, chronic obstructive pulmonary disease and renal failure), which have shown prognostic significance in previous studies.

The study by Badano et al. represents a significant advance in the understanding and management of STR, offering a novel approach to risk stratification and a potential tool for guiding treatment decisions. The integration of advanced cardiac imaging with sophisticated data analysis and machine learning exemplifies the future of personalized cardiology. As these methods are further refined and validated, we can anticipate a paradigm shift in the management of STR and other complex cardiovascular conditions. The transition from the stethoscope to artificial intelligence is more than a technological leap—it is a transformative step in how we understand and treat cardiovascular disease.

The current challenge lies in translating these findings into practical tools for daily clinical use, ensuring that the benefits of this innovative research reach the patients who need them most. This study marks a meaningful advance toward achieving a more nuanced understanding and individualized approach to STR.