GATIS score: An innovative prognostic score for rectal neuroendocrine neoplasms

Abstract

In this article, we discussed the article by Zeng et al, published in a recent issue of the World Journal of Gastroenterology. The publication represents a significant advancement in the prognostic evaluation of rectal neuroendocrine neoplasms. The GATIS score is a single nomogram model that incorporates five key prognostic factors: Tumor grade; T stage; tumor size; age; and the prognostic nutritional index. This innovation optimizes the prognostic process, delivering more accurate predictions of overall survival and progression-free survival compared to traditional TNM staging and World Health Organization classification systems. The findings of the study were based on a retrospective analysis spanning 12 years and involving 1408 patients from 17 reference centers in China. In this editorial, we specifically examined the strengths and limitations of the study, the clinical implications of the GATIS score, and the questions arising from its conclusions.

Key Words: Rectal neuroendocrine neoplasm; Prognosis; Score; Overall survival; Progression-free survival

Core Tip: The GATIS score enables a more robust prognostic evaluation of patients with rectal neuroendocrine neoplasms. This prognostic score evaluates key factors, such as tumor grade, T stage, tumor size, age, and the prognostic nutritional index, and surpasses traditional models in predicting overall and progression-free survival. This advancement highlights the value of personalized medicine and standardized prognostic evaluations in enhancing patient outcomes and clinical practice.

TO THE EDITOR

Rectal neuroendocrine neoplasms (R-NENs) are a heterogeneous group of malignancies that arise in the rectum and originate from neuroendocrine cells[1,2]. Although rare, the prevalence of these malignancies is increasing, underscoring the need for both optimal management and accurate prognostication for affected patients[3]. These tumors are generally associated with a favorable prognosis, with a 5-year survival rate reaching as high as 90%[4]. Numerous factors influence patient outcomes, including tumor size, tumor grade, depth of invasion, lymphovascular invasion, lymph node involvement, metastases, differentiation, hormonal activity, and additional variables such as the patient’s overall health, age, and comorbidities[5,6]. Consequently, a reliable prognostic score is essential to effectively predict both overall survival (OS) and progression-free survival (PFS).

The traditional World Health Organization (WHO) classification and TNM staging systems attempt to address the need for a reliable and valid prognostic score[7,8]. WHO classification separates tumors based on the morphology and malignant potential of the tumor cells, while TNM classifies them according to the size of the primary tumor, lymph node involvement, and the presence or absence of distant metastasis. Although these systems are reliable, they have several limitations in their predictive accuracy. This underscores the need for a more comprehensive and robust prognostic model.

The recent publication by Zeng et al[9] in the World Journal of Gastroenterology marks a significant advancement in the prognostic evaluation of R-NENs. The study, titled “GATIS score for predicting the prognosis of rectal neuroendocrine neoplasms: A Chinese multicenter study of 12-year experience,” introduced the development and validation of the GATIS score, a novel predictive model. This model was based on a thorough analysis of a large cohort comprising 1408 patients from 17 referral medical centers in China. In this study, key prognostic factors were identified. The GATIS score incorporates five critical indicators: Tumor grade; T stage; tumor size; age; and the prognostic nutritional index (PNI).

This article aimed to underscore the importance and potential impact of the GATIS score as described in the study by Zeng et al[9]. Specifically, we highlighted the advancements this model brings to oncology and the clinical implications it offers. Additionally, this editorial explored the strengths and limitations of the study and addressed the questions arising from its findings.

Critical analysis

The study by Zeng et al[9] had numerous strengths, which significantly increased the reliability of the study. The primary strength was the rigorous methodological framework, which increased the robustness of the predictive model. In addition, this study used a relatively large sample size especially given the rarity of the disease. This large cohort improved the reliability of the findings while supporting their generalizability. Data collection from 17 referral medical centers over 12 years was another strength of the study and increased the validity of the GATIS score as it ensures that the findings are not limited to the practices or patient population of a single institution.

The utilization of multiple prognostic factors, including tumor grade, T stage, tumor size, age, and the PNI provides more accurate predictions of patient outcomes compared to models that rely on fewer variables. Additionally, the use of decision curve analysis to evaluate the clinical utility of the GATIS score was another significant strength, as it quantifies the benefit of implementing the score in clinical practice.

Another strength of the study was its high predictive accuracy. The results demonstrated high C-index values for OS (0.915) and PFS (0.908), reflecting strong predictive performance of the GATIS score. Specifically, this performance significantly surpassed that of the traditional WHO classification and TNM staging systems, typically exhibiting C-index values below 0.8. Notably, most existing predictive models in the literature, including machine learning approaches, fail to achieve comparable accuracy. Specifically, there are several predictive models available in the literature, with most of them being machine learning[10-13]. For example, while a specific nomogram achieves an area under the receiver operating characteristic curve of 0.937 for predicting lymph node metastasis (a critical prognostic factor), models based on a single variable remain inherently limited[11]. The prognostic models, which utilize data from the SEER database, designed the nomogram based on more than 10000 patients, which contributes to their reliability and the generalizability of the results. These findings suggest that subsequent studies on the GATIS score could benefit from even larger patient cohorts. Nonetheless, machine learning predictive models do not provide a predictive scoring system.

Despite its strengths, the study also had notable limitations. First, the retrospective design introduced risks of selection and information bias. Another limitation of retrospective studies is the possible lack of patient data that could potentially affect the outcome of the statistical analysis. Furthermore, the absence of long-term follow-up of the patients, a key feature for the design of a prognostic model, limits the validity of the GATIS score. Finally, the study focused primarily on the clinical and pathological anatomical picture of the patients, without considering molecular and genetic data. This omission significantly limits the validity of the GATIS score, leaving room for future enhancements.

Another limitation of the study was the non-inclusion of the latest data available in the literature. For instance, somatostatin receptor 2 is expressed in more than two-thirds of patients with rectal neuroendocrine tumors. These patients are characterized by a favorable prognosis and good OS. Therefore, the expression of somatostatin receptor 2 could be used as an additional prognostic factor[14]. Similarly, the type of surgery impacts patient outcomes, with endoscopic submucosal dissection and endoscopic mucosal resection showing favorable prognoses for rectal neuroendocrine tumors, even in cases with positive resection margins[15]. Other variables influencing the prognosis include sex, race, alkaline phosphatase, the presence of lymph node metastases, and the number of affected lymph nodes[16-20]. Finally, certain endoscopic findings predispose patients to an increased risk of recurrence. These include female sex, redness, G2 tumor grade, and non- endoscopic submucosal dissection methods[21].

Clinical implications

Undoubtedly, the most important clinical implication of the GATIS score is the improved prognostic accuracy compared to the other available prognostic scores. This improvement allows clinicians to make more accurate prognostic assessments, which are critical for counselling patients and setting realistic expectations about disease progression. Specifically, this improvement is found in both the OS and PFS. This increased score performance also contributes to the adaptation of neoplasm management strategies, balancing the potential benefits and risks of various interventions. For patients with localized R-NENs, the GATIS score can also contribute to the correct selection of surgical interventions, both in terms of their type and timing. By incorporating multiple prognostic factors, the GATIS score allows a more detailed categorization of patients based on their individual risk profiles. Patients with a high risk can be identified for more aggressive treatments. On the contrary, patients with a low risk may benefit from less aggressive treatments, thus avoiding unnecessary operations that also include a risk of complications. The health system also benefits from this through the reduction of costs, thus offering a better allocation of resources.

In addition, by integrating the PNI index into the prognostic model, the nutritional and immune status of a patient is also considered. Therefore, this index offers a more holistic approach to patients, unlike the traditional TNM and WHO classifications.

Using the GATIS score can also improve communication with patients. More specifically, the use of a reliable prognostic score allows patients to understand their prognosis. This understanding leads to the ability to make more conscious decisions. In addition, patients with a poor prognosis can be given supportive care and end-of-life care. Finally, with an understanding of the prognosis, and under appropriate support, patients can manage psychological stress more effectively.

The contribution of a reliable prognostic score to the field of research is also important. Specifically, with a more reliable prognostic score, patients with a high risk can benefit from experimental treatments, and researchers can target interventions more effectively. The GATIS score can be used to better categorize patients in clinical trials, ensuring that participants have similar prognostic characteristics. This can increase the validity of clinical trial results. Furthermore, the success of the GATIS score in R-NENs may inspire the development of similar prognostic models for other types of cancer. The GATIS score may also play a role in the education and training of oncology professionals.

Future directions

While the development of the GATIS score constitutes a significant advancement in the prognostic evaluation of R-NENs, many questions were raised that require further investigations. Initially, prospective validation studies are required to confirm the results of this study. These studies should be conducted across diverse regions to ensure the broader applicability of the model. Additionally, integrating long-term follow-up data is crucial for validating the score and establishing its utility in clinical practice. The collection of these data is fundamental to the widespread adoption of the GATIS score.

Future research should also explore incorporating molecular and genetic biomarkers into the GATIS score to enhance its predictive power. The lack of such data is a major limitation of the current study. Addressing this gap would make the nomogram more robust and reliable. Also, studies can be conducted to investigate how the GATIS score correlates with treatment responses over time. The integration of patient reported outcomes can further help in designing a more holistic score that will also analyze patient quality of life. Also, to increase the performance of the prognostic score, it is worth investigating the integration of the remaining factors that may influence the prognosis.

Educational and training programs for healthcare professionals are necessary to ensure familiarity with the GATIS score and its application in clinical practice. Thus, further information can be obtained on the value of the score in daily clinical practice. The design of automated tools and software applications can also contribute and will significantly increase the ease of use of the score and its usefulness.

Finally, it is worth investigating the development and validation of similar prognostic models for other types of neuroendocrine tumors and cancers. Given the high performance of the GATIS score, new scores can significantly contribute to the reliable prognostic assessment of patients.

CONCLUSION

The GATIS score is a significant marker for the prognosis of patients with R-NENs, offering numerous clinical applications. While it demonstrated considerable strengths, it also raised important questions, which will further increase the prognostic evaluation of these patients. The study by Zeng et al[9], which introduced this nomogram, is characterized by many strengths that validated the results. However, addressing the limitations of the study in future research will be essential to refine the score and facilitate its integration into clinical practice.