A new score for predicting incidental appendiceal neoplasms in patients aged ≥40 years with acute appendicitis: a multicenter retrospective cohort study.

Acute appendicitis, defined as inflammation of the vermiform appendix, is a prevalent surgical emergency worldwide, with a lifetime risk of approximately 8.6% in men and 6.7% in women. [ 1 ] The global incidence of the condition ranges from 86 to 160 cases per 100,000 population per year, depending on the region. [ 2 ] Although benign pathologies are typically observed in this common surgical procedure, clinically significant incidental appendiceal neoplasms may be found in the tissue removed during appendectomy. The reported incidence of incidental neoplasia ranges from 0.2% to 2.5% across various case studies. [ 3 , 4 ] Recent studies suggest that the incidence is higher in older populations, highlighting the need for improved preoperative risk stratification. The most prevalent histologic types are neuroendocrine tumors (NETs) and low-grade mucinous neoplasms. [ 4 ] These lesions are typically clinically silent preoperatively and often difficult to identify radiologically. [ 5 ] The presence of unidentified incidental tumors may adversely affect prognosis due to inadequate resection and incomplete staging. Recent studies have shown that right hemicolectomy may offer a survival advantage over appendectomy, particularly for T2–T3 stage appendiceal tumors. [ 6 ] Additionally, the National Comprehensive Cancer Network (NCCN) guidelines recommend right hemicolectomy with at least 12 lymph node dissections for adenocarcinoma and goblet cell types. [ 7 ] These findings underscore the importance of stage-specific surgical planning for incidental tumors. The development of several predictors has been identified as a means to enhance preoperative diagnostic accuracy. Increased appendix diameter, thickening of the appendiceal wall, and calcification of the appendix on computed tomography (CT) may raise suspicion of neoplasia. [ 5 ] Clinical parameters, including advanced age and lower white blood cell counts, have also been associated with increased risk. [ 8 ] In the study by Er et al., [ 9 ] three parameters were identified as predictive factors: appendiceal diameter above 11.5 mm, absence of wall contrast, and low neutrophil count. However, due to the use of propensity score matching, factors such as age and gender were not incorporated into the study design. Notably, multiple studies have reported a clear increase in the risk of neoplasia after the age of 40, suggesting that both age and gender may have an important influence on predictive modeling. [ 10 ] This multicenter retrospective cohort study was conducted to address this gap. The primary aim was to develop and validate a novel five-parameter scoring system that integrates age, gender, appendix diameter on CT, absence of wall contrast enhancement, and neutrophil count to improve preoperative prediction of incidental appendiceal neoplasms in patients aged ≥40 years. We hypothesized that incorporating demographic parameters with clinical and radiologic findings would enhance predictive performance compared to existing models. Given the prognostic implications of undiagnosed neoplasms, an effective risk stratification tool could better guide surgical planning and postoperative management. This scoring system may be useful in the clinical decision-making process during the preoperative period for risk classification and surgical planning. It may also be valuable intraoperatively for determining the surgical approach in high-risk patients based on frozen section evaluation, and postoperatively for monitoring pathological results and prioritizing colonoscopy and oncologic follow-up. This system was developed as a complementary tool to existing appendicitis diagnostic scoring systems, not as an independent diagnostic method intended to replace them.

After applying the inclusion and exclusion criteria, a total of 2,143 patients aged 40 years and older were included in the final analysis. The cohort consisted of 952 female patients (44.4%) and 1,191 male patients (55.6%), with a mean age of 52.64±10.70 years. All patients underwent appendectomy for suspected acute appendicitis, and histopathologic examination was performed for every specimen collected. Histopathologic evaluation revealed incidental appendiceal neoplasia in 122 patients (5.7%), with the detailed distribution of diagnoses presented in Table 1 . Comparative analysis demonstrated a statistically significant difference in age between patients with neoplasia and those with benign findings (57.29±13.17 years vs. 52.36±10.47 years, p<0.001). The proportion of female patients was significantly higher in the neoplasia group compared to the benign group (7.1% vs. 4.5%; p=0.010). Neutrophil counts were significantly lower among patients diagnosed with neoplasia (8.27±3.72×10 9 /L vs. 10.01±4.03×10 9 /L, p<0.001). Furthermore, the mean appendix diameter was larger in patients with neoplasia (12.19±6.76 mm versus 10.48±2.75 mm, p<0.001). The absence of contrast enhancement on computed tomography was also observed more frequently in the neoplasia group (p=0.009). Detailed characteristics and comparative data of the study population are provided in Table 3 . Patients’ data Univariate and multivariate logistic regression analyses identified age, neutrophil count, appendix diameter, and absence of contrast enhancement as independent predictors of neoplasia. Female gender showed borderline statistical significance in the multivariate analysis (odds ratio: 1.44; p=0.058). The β coefficients, odds ratios, confidence intervals, and p-values are summarized in Table 4 . Univariate and multivariate regression analysis OR: Odds Ratio; CI: Confidence Interval. The diagnostic performance of the five-parameter scoring system was assessed using receiver operating characteristic analysis. The area under the curve was 0.641 (95% confidence interval: 0.594–0.690; p<0.01). When a threshold score of three or higher was used to predict neoplasia, the sensitivity was 56.6% and the specificity was 65.4%. The positive predictive value was 9%, and the negative predictive value was 96.1%. The positive likelihood ratio was 1.63, and the negative likelihood ratio was 0.66. Notably, patients who achieved the maximum score of 5 had the highest observed incidence of neoplasia, reaching 33.3%. The distribution of neoplasia rates according to scoring categories is shown in Table 5 . Effectiveness of scoring in the patient group In addition to sensitivity, specificity, PPV, and NPV, a confusion matrix was constructed for the primary threshold (≥3 points) to provide a clearer representation of diagnostic accuracy. The distribution of true positives, false positives, true negatives, and false negatives is shown in Table 6 . This matrix highlights the relatively high negative predictive value of the score, with most patients classified as low risk (<3 points) correctly identified as benign. Confusion matrix for cut-off ≥3 TP (True Positive): Patients with neoplasia correctly identified by the score (≥3). FP (False Positive): Patients without neoplasia incorrectly identified as high-risk (≥3). FN (False Negative): Patients with neoplasia incorrectly classified as low-risk (<3). TN (True Negative): Patients without neoplasia correctly identified as low-risk (<3).

This multicenter retrospective cohort study presented a new five-parameter score, comprising age, sex, appendix diameter on CT, absence of wall contrast enhancement, and neutrophil count, to predict incidental appendiceal neoplasms in patients aged ≥40 years with acute appendicitis. The model’s discriminatory power (AUC≈0.64), and particularly its high negative predictive value (≈95–96%), allowed for confident discrimination among low-risk patients. The incidence was significantly higher in the high-risk group (≈33% at the highest score level of 5 points). This indicates that the proposed model may serve as a reliable adjunct to stratify patients into low- and high-risk groups, supporting integrated risk assessment with imaging and pathology rather than clinical decision-making based on the score alone. Each parameter in the model was assigned one point. This decision was made because weighting based on statistical coefficients would only slightly increase the area under the curve but would complicate clinical application. Equal scoring provides an approach that is simple for clinicians to use in daily practice. Therefore, our model aims to balance performance and applicability. The key strength of this study is its reliance on a multicenter, large patient series. This allowed the development of a new scoring model that integrates both clinical and radiological parameters, focusing on patients aged ≥40 years. While previous scores have primarily focused on identifying the inflammatory process, our study is one of the first large-scale attempts to predict the risk of incidental neoplasia. Furthermore, the inclusion of the parameter absence of wall enhancement represents an innovative contribution to the literature. Clinically, this scoring system may provide additional decision support in patients aged ≥40 years undergoing appendectomy, particularly by informing the extent of resection and the need for thorough histopathological follow-up. In the literature, large series have demonstrated that advanced age and increased appendix diameter are associated with malignancy. In particular, the combination of age over 40 and an appendix diameter greater than 10 mm on CT has been shown to increase the risk of malignancy more than threefold. [ 11 ] Similarly, large databases have reported that the likelihood of appendiceal adenocarcinoma increases with age, with most adenocarcinomas occurring around age 40. [ 14 ] In our series, the risk of malignancy was also found to increase in individuals over 40 years (AUC=0.732). This supports the view that predicting malignancy risk is clinically important, especially in this age group. Various clinical scoring systems have long been used in the diagnosis of acute appendicitis. The Alvarado score was the first system based on symptoms, physical examination findings, and laboratory values, and is still widely used. [ 15 ] The Appendicitis Inflammatory Response (AIR) score and the Adult Appendicitis Score (AAS), developed subsequently, aimed to improve diagnostic accuracy and reduce negative appendectomy rates. [ 16 , 17 ] The World Society of Emergency Surgery (WSES) 2020 guideline also stated that AIR and AAS are the clinical scores with the highest discriminatory power in adults. [ 18 ] However, all of these scores are designed solely to distinguish inflammatory processes and lack the ability to predict incidental appendiceal neoplasms. Our proposed score differs by specifically addressing neoplasia prediction rather than inflammation, thereby complementing existing clinical tools. While AIR and AAS have demonstrated high diagnostic accuracy in differentiating uncomplicated from complicated appendicitis, [ 16 ] none of these systems include oncologic endpoints. Thus, our score should be considered complementary, providing oncologic risk stratification that existing clinical scores do not address. Attempts to develop a scoring system to predict appendiceal neoplasia are quite limited in the literature. Although contrast enhancement of the appendiceal wall has been demonstrated as a valuable diagnostic feature in acute appendicitis—with sensitivity around 75% and specificity around 85% on contrast-enhanced CT [ 19 ] —this parameter has not previously been applied to neoplasia prediction. To our knowledge, Er et al. [ 9 ] were the first to report the absence of wall enhancement as a factor associated with incidental neoplasia. This study builds on that evidence and makes an innovative contribution by incorporating the absence of wall enhancement into a validated neoplasia risk score for the first time in the literature. In this study, the analysis population was limited to patients aged 40 years and older, who are considered to have a significantly increased risk of incidental appendiceal neoplasia in the literature. [ 11 - 13 ] However, subgroup ROC analysis demonstrated that age 50 represents an additional risk threshold (AUC=0.605); therefore, age 50 was included in the scoring system as ≥1 point. Thus, the model not only targeted the population aged ≥40 years but also reflected the gradual increase in risk within this group. This suggests that age may act as a stepwise rather than linear risk factor, enabling more refined risk stratification in clinical practice. To enhance the clinical applicability of our model, we determined how to interpret the proposed score across different patient subgroups. Patients scoring 0-2 points (low risk) have a very low probability of incidental neoplasia (NPV >95%), and standard appendectomy with routine follow-up appears sufficient for these patients. Patients scoring 3-4 points (intermediate risk) may be considered for intraoperative frozen section, if technically feasible, and should be prioritized for postoperative colonoscopy and closer oncological follow-up. In contrast, patients scoring 5 points (high risk) demonstrate a neoplasia probability of approximately one-third, and intraoperative frozen section or even extended resection (e.g., right hemicolectomy), along with close oncological surveillance, may be necessary for these patients. However, this study has some limitations. The retrospective design increases the risk of selection bias, and potential differences in CT protocols between centers may have led to standardization issues, particularly regarding wall enhancement assessment. Nonetheless, it should be noted that this study was conducted at tertiary centers, with imaging evaluated by experienced radiologists. Furthermore, the model's discriminatory power was found to be moderate (AUC≈0.64); therefore, it should be interpreted as an adjunct tool rather than a standalone clinical decision-maker. Another limitation is the lack of prospective validation, which will be essential to confirm reproducibility across different healthcare settings and geographic regions. Another limitation of the study was that IT protocols and pathology assessments were not fully standardized between centers. However, common definitions were used to reduce inter-observer variability. In summary, this novel five-parameter score represents an innovative step toward predicting incidental appendiceal neoplasia in patients aged ≥40 years with acute appendicitis. By combining age, sex, CT findings, and laboratory parameters, it may guide clinicians in tailoring surgical strategies, optimizing pathology evaluation, and avoiding unnecessary interventions in low-risk patients. Future studies should prospectively validate this model and assess its integration with molecular biomarkers and artificial intelligence (AI)-assisted imaging to further enhance risk prediction.

This multicenter retrospective cohort study introduced a novel five-parameter score—including age, sex, appendix diameter, absence of wall contrast enhancement, and neutrophil count—to predict incidental appendiceal neoplasms in patients aged ≥40 years with acute appendicitis. The model showed moderate discriminatory power (AUC≈0.64) but a high negative predictive value (≈95–96%), supporting its use as an adjunct for risk stratification. Unlike traditional appendicitis scores, this tool specifically targets neoplasia prediction and may guide surgical decision-making by identifying high- and low-risk patients. Prospective validation and integration with molecular or AI-based methods are needed to enhance predictive performance.

This was a multicenter retrospective cohort study involving six tertiary healthcare institutions. Patients who underwent appendectomy with a preliminary diagnosis of acute appendicitis between January 2019 and December 2024 were retrospectively screened using electronic patient record systems. All participating centers followed a standardized protocol for data extraction and analysis. At the onset of the study, a total of 8,110 patients aged ≥18 years underwent retrospective evaluation. The final analysis included 6,130 patients who met the predetermined inclusion and exclusion criteria. Preliminary evaluation of the entire population showed that age was the strongest predictor of neoplasia. The area under the curve (AUC) obtained for age in the receiver operating characteristic (ROC) analysis was 0.732 (95% confidence interval [CI]: 0.691-0.773). The cut-off value of 40 years yielded a sensitivity of 68.5% and a specificity of 70.2%. Additionally, the literature has shown that the risk of appendiceal neoplasia varies significantly after the age of 40. [ 11 - 13 ] Based on these findings, further analysis was limited to a subgroup of patients aged over 40 years to increase clinical relevance. The study flow diagram is presented in Figure 1 , summarizing patient selection, exclusions, and final cohort composition. Patient selection flow diagram. This study was retrospective in design, and no power analysis was performed beforehand to determine sample size. All cases meeting the specified criteria were included in the analysis. This study was approved by the institutional review boards of all participating centers (Approval No: TABED-2-25-1329). Informed consent was waived due to the retrospective design. The study was conducted in accordance with the principles of the Declaration of Helsinki. The inclusion criteria were as follows: patients aged 40 years or older who underwent appendectomy for acute appendicitis, received preoperative contrast-enhanced abdominal computed tomography (CT), had preoperative neutrophil levels recorded in the laboratory system, and had histopathologic examination results available. The exclusion criteria were: pregnant women; patients with a known history of malignancy; those who underwent incidental appendectomy during other abdominal surgery (e.g., gynecologic procedures); and patients with incomplete or inconsistent data records. Data were obtained from the hospital’s automated systems and patient files. The following variables were recorded at all centers using a standardized data collection form: Age (years), Gender (female, male), Diameter of the appendix (mm; widest transverse measurement on CT), Wall contrast uptake on CT (present, absent) (the absence of wall contrast was defined as a lack of mural contrast retention compared to the adjacent bowel wall. Due to the multicenter design, minor differences existed in CT scanning protocols and pathology reporting criteria; this was noted as a study limitation), Preoperative neutrophil count (10 9 /L), Histopathologic diagnosis (binary coding as neoplasia/benign). Age (years), Gender (female, male), Diameter of the appendix (mm; widest transverse measurement on CT), Wall contrast uptake on CT (present, absent) (the absence of wall contrast was defined as a lack of mural contrast retention compared to the adjacent bowel wall. Due to the multicenter design, minor differences existed in CT scanning protocols and pathology reporting criteria; this was noted as a study limitation), Preoperative neutrophil count (10 9 /L), Histopathologic diagnosis (binary coding as neoplasia/benign). Histopathologic diagnoses were classified into benign and neoplasia categories based on established oncologic criteria ( Table 1 ). Cases with missing critical variables were excluded from the analysis. Distribution of patients according to histopathologic diagnoses A scoring system was proposed based on age, female gender, appendix diameter, absence of wall contrast enhancement on tomography, and low neutrophil count ( Table 2 ). The total score ranged from 0 to 5. Scoring system ROC: Receiver Operating Characteristic; AUC: Area Under the Curve; CI: Confidence Interval. Data were analyzed using IBM SPSS Statistics for Windows, Version 22.0 (IBM Corp., Armonk, NY, USA). Continuous variables were expressed as mean±standard deviation, and categorical variables were expressed as frequency and percentage. The diagnostic performance of the scoring system was evaluated using ROC analysis. The AUC, 95% confidence interval, and p-value were calculated. In addition, diagnostic performance measures such as sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood Ratio (LR+), and negative likelihood ratio (LR-) were reported for different threshold values. Univariate and multivariate logistic regression analyses were conducted to identify independent predictors. Missing data were managed through case-wise deletion. Statistical significance was set at p<0.05.