Enteropathy-associated T-cell lymphoma (EATL) is a rare and aggressive T-Cell Non-Hodgkin lymphoma that primarily arises in the small intestine and is closely associated with celiac disease (1–4). The incidence of EATL remains low, making it a challenging entity to study. The prognosis of EATL is generally poor, with reported median 5-year overall survival (OS) ranging from 0% to 59% (5). In patients with EATL who have a prior diagnosis of refractory celiac disease, the 5-year OS is significantly lower, ranging from 0% to 8% (5–9). Treatment for this disease typically involves surgery and chemotherapy, but the prognosis remains discouraging, with OS rates being dismal, particularly in advanced stages (1, 10, 11). Given the rarity of this malignancy, large-scale studies to fully understand its epidemiology, prognostic factors, and long-term survival outcomes remain sparse. Therefore, comprehensive studies of EATL are crucial to improve clinical management and patient outcomes.

Despite the recognition of EATL’s aggressive nature, studies investigating its survival prognosis have been limited, and available data are mostly derived from case reports or small series. Also, current survival models often fail to address the dynamic nature of survival outcomes, particularly in terms of changing prognostic factors over time. In this study, we compare the advantages of conditional survival (CS) analysis to traditional survival analysis, which allows for a more nuanced view of survival. CS analysis can identify critical time points post-diagnosis where mortality risk is highest, thus offering valuable insights into the timing of intervention and patient management (12–15).

To address these gaps in research, we leveraged the SEER database, which provides a large cohort with long-term follow-up data. The SEER database is invaluable for studying rare cancers like EATL, as it offers a robust sample size that allows for more reliable analysis of epidemiological trends, survival outcomes, and the development of novel prognostic models. In this study, we aim to evaluate the incidence trends and use CS analysis to better understand the epidemiological information and survival prognosis of EATL patients. Additionally, we developed a novel nomogram model based on these dynamic survival factors to provide more accurate outcome predictions.

EATL is a rare and aggressive form of T-cell non-Hodgkin lymphoma that primarily arises in the small intestine and is strongly associated with celiac disease (3, 5, 10). Due to its extremely low incidence, the study of EATL presents significant challenges, particularly in terms of obtaining large-scale data for comprehensive analysis. Given the poor prognosis of EATL, it is essential to conduct large-scale studies with long-term follow-up to better understand its incidence trends, prognostic factors, and survival outcomes. In this study, we attempted to fill these gaps by analyzing the incidence trends of EATL, identifying significant prognostic factors, and using CS analysis to develop a dynamic prognostic model. Through these analyses, we have gained valuable insights into the epidemiological characteristics and survival prognosis of EATL, offering a more accurate and clinically relevant tool for managing this challenging malignancy.

The incidence rate of EATL we calculated was 0.014 per 100,000. Two prior population-based studies from the United States reported incidence rates of 0.011 and 0.016 per 100,000 per year, while population-based studies from Europe reported rates of 0.04, 0.07, and 0.1 per 100,000 per year (10, 19–21). The analysis of EATL’s incidence trends revealed a concerning upward trajectory over the past two decades, with an APC of 2.63. Although EATL remains a rare disease, this increasing incidence could reflect greater awareness, improved diagnostic methods, or potentially an actual rise in cases. The extremely low incidence of EATL makes it difficult to track accurately, and this upward trend may indicate a broader recognition of the disease in clinical practice. The reasons for the increase in incidence remain speculative, but factors such as better reporting, advancements in diagnostic technologies (e.g., molecular techniques), and the growing recognition of its link to celiac disease may contribute to the observed trend. However, further investigation with large cohort studies is needed to establish the underlying causes definitively.

Our study identified several significant prognostic factors for OS in EATL patients, including age, tumor site, tumor site, tumor stage, surgery, and chemotherapy. Age and tumor stage were identified as independent risk factors, with older age and advanced tumor stage associated with poorer survival outcomes, consistent with previous studies (5, 8, 20, 22). The role of surgery and chemotherapy in prolonging survival highlighted the necessity of timely intervention and appropriate therapeutic strategies. And several studies have also highlighted the clinical efficacy of combining surgery with chemotherapy, which significantly improved the clinical outcomes of these tumors (8, 10, 23). However, despite treatment, the prognosis for EATL remains extremely poor, as reflected in the low 5-year survival rates. This reinforced the need for improved treatment protocols and the exploration of novel therapeutic strategies to address this aggressive lymphoma.

Given the dismal prognosis of EATL, we further explored its survival outcomes using CS analysis. This analysis allowed us to assess how survival probabilities change over time, providing a more nuanced understanding of the disease’s progression. Our findings revealed that the highest mortality rate occurs within the first year post-diagnosis, with survival probabilities significantly improving after surviving this critical period. This underscored the importance of early detection and intervention in EATL. The dynamic nature of CS analysis, which accounts for changing prognostic factors over time, offers valuable insights into the most critical time points for patient management. By identifying these high-risk periods, clinicians can better tailor treatment plans to improve survival chances during the most vulnerable phases of the disease.

The development of a nomogram model based on CS analysis represents a significant advancement in predicting survival outcomes for EATL patients (24). By integrating dynamic survival probabilities at different time points, this model provided a more accurate prediction of individual patient survival. The ability to stratify patients into low- and high-risk groups based on their predicted survival probabilities can guide clinical decision-making and help prioritize treatment interventions. This model’s clinical utility is particularly valuable, as it combines traditional prognostic factors with dynamic survival data, offering a comprehensive approach to managing EATL. The integration of CS analysis with the nomogram enhanced its accuracy and provided clinicians with a tool that reflected not only initial prognosis but also evolving survival probabilities over time (25–28).

Despite the strengths of our study, several limitations inherent to the SEER database should be acknowledged. First, in 2016 the WHO classification redefined monomorphic epitheliotropic intestinal T-cell lymphoma (MEITL) as a distinct entity, previously grouped with EATL. Because SEER lacks comprehensive histopathological details, particularly for cases diagnosed before 2016, a complete separation of these two entities was not feasible. Consequently, some MEITL cases may have been misclassified as EATL, potentially influencing our incidence and survival estimates. Second, the SEER database does not provide tumor-level biological information, including recurrent molecular alterations (SETD2, STAT5B, TP53), T-cell receptor subtype (γ/δ vs. α/β), or immunophenotypic markers such as CD8 and CD56. These features are increasingly recognized as clinically relevant for disease classification, treatment response, and prognosis in EATL/MEITL, but they could not be incorporated into our analysis or nomogram. Therefore, our findings are based mainly on demographic and clinicopathological characteristics. Future multi-institutional studies that integrate detailed pathological, immunophenotypic, and genomic data with outcomes are needed to refine risk stratification and develop biologically informed prognostic tools. Third, SEER lacks information on treatment details, including the timing and sequence of surgery (at diagnosis vs. later in the disease course, before vs. after chemotherapy) and chemotherapy regimen heterogeneity (specific agents, dosing schedules, or subsequent lines of therapy). This limited granularity prevents assessment of the prognostic impact of surgical timing, comparisons across chemotherapy regimens, or outcomes in patients who fail frontline therapy. Given the clinical relevance of these factors, future prospective or multi-institutional retrospective datasets with detailed therapeutic information will be critical for clarifying their influence on survival. In addition, SEER does not capture information on the presence of celiac disease or refractory celiac disease, which restricts our ability to explore the observed increase in EATL incidence. Finally, as a retrospective registry-based study, our findings are subject to potential biases and may not be generalizable to all populations. External validation of our model in independent cohorts and prospective studies will be essential to strengthen its clinical applicability.

Looking ahead, our findings lay the groundwork for the development of clinical decision-support tools to aid in the management of EATL. The CS–integrated nomogram we constructed demonstrated the feasibility of combining weighted clinicopathological variables with dynamic survival probabilities to generate individualized prognostic estimates. Such tools could be further refined by incorporating molecular alterations, immunophenotypic markers, treatment sequence, and regimen-specific details as these data become available in future multi-institutional datasets. Ultimately, integrating these enriched models into web-based platforms or electronic health record systems may provide clinicians with real-time, patient-specific prognostic information, thereby supporting risk-adapted treatment strategies, optimizing resource allocation, and improving patient counseling.

Our study, utilizing the SEER database, provided valuable insights into the incidence trends, prognostic factors, and survival outcomes of EATL. The increasing incidence of EATL over the past two decades highlighted the growing recognition of this malignancy. We identified several independent prognostic factors, such as age, tumor site, tumor stage, surgery, and chemotherapy. Through CS analysis, we provided a more dynamic and nuanced understanding of the disease’s progression, demonstrating the critical importance of early intervention, particularly in the first year post-diagnosis. The development of a novel nomogram model, integrating dynamic survival data, offers clinicians a powerful tool for predicting individual survival outcomes and improving patient management.