Appendiceal adenocarcinoma (AA) is a rare tumor of the gastrointestinal tract.1,2 Consensus guidelines for the management of appendiceal adenocarcinoma follows that of colorectal cancer (CRC), given the scarcity of high-quality data and their presumed clinical similarity arising from their common embryologic origin and anatomic proximity.3,4,5 Surgical management of nonmetastatic AA includes appendectomy or right hemicolectomy. Appendectomy alone is currently recommended for patients with intramucosal AA or carcinoma invading the submucosa without invasion of the muscularis propria (T1) and no high-risk features (i.e., low grade, negative resection margins, no lymphovascular invasion). Patients with high-risk T1 tumors or invasion beyond the submucosa (> T2) are considered for a right hemicolectomy for pathologic assessment of the ileocolic lymph nodes.6,7

The clinical behavior of appendiceal adenocarcinomas varies on the basis of disease stage and the presence of a mucinous subtype.8 Mucinous appendiceal adenocarcinoma (MAA) is distinguished from non-mucinous appendiceal adenocarcinoma (NMAA) by the criterion of having > 50% of the cross-sectional area histologically comprised of mucin.3 MAA have distinct patterns of metastatic spread, with mucinous seeding of the peritoneal cavity and a lower rate of lymphatic and hematogenous dissemination.9,10 While the rate of isolated lymph node metastases in the absence of other metastatic disease is lower for MAA compared with NMAA, the presence of nodal metastases is nevertheless prognostic of survival for both subtypes.11,12 Still, the role of routine colectomy for therapeutic removal of involved lymph nodes may not improve survival, particularly for MAAs. Given the lack of randomized trial data, the appropriate oncologic extent of resection for localized AAs has been guided by several cancer-registry-based analyses, some of which have grouped together both mucinous and non-mucinous histologies, or have been limited by use of overall survival as a study endpoint—where patient comorbidity may bias both extent of resection and the rate of non-disease-related deaths.13,14,15,16 To clarify the association between extent of resection and disease-specific survival (DSS), a preferrable study endpoint for a cancer with a generally favorably prognosis, the Surveillance, Epidemiology, and End Results (SEER) database was queried to evaluate the oncologic value of the extent of surgical resection for both mucinous and non-mucinous subtypes.

Patients and Methods

Patient selection

After institutional review board approval, data (2000–2020) were extracted from the SEER database (https://seer.cancer.gov/). Patients ≥ 18 years of age who underwent resection of appendiceal adenocarcinoma were identified. Patients were selected using appendix site code (ICD-O-3 topography code C18.1) and corresponding ICD-O-3 morphology codes for MAA and NMAA (Supplemental Table 1). Patients were excluded if they had metastatic disease at the time of surgical resection or unknown disease-specific survival.

Variables

The demographic and clinical SEER variables utilized in this study included age (< 60, 60–69, 70–79, and > 80 years), sex, race/ethnicity, median income, population density (metropolitan areas, adjacent or non-adjacent nonmetropolitan areas), primary tumor (pT)-stage, regional lymph node metastasis (pN)-stage, tumor grade, and administration of adjuvant chemotherapy. Extent of resection was defined by either appendectomy versus hemicolectomy. The primary outcome, DSS, was defined as the interval between date of diagnosis and date of death from disease, with censoring at last contact or death from another cause.

Statistical Analyses

Descriptive statistics are presented as frequencies for categorical variables and median [interquartile range (IQR)] for continuous variables. Pearson’s χ2 and Wilcoxon rank-sum tests were used to analyze categorical and continuous variables, respectively. Variables associated with colectomy on univariate analysis were entered into a stepwise logistic regression model (p ≤ 0.05 for entry; p > 0.10 for removal) to identify independent predictors of colectomy utilization. The influence of colectomy on DSS was analyzed using Kaplan–Meier estimates and Cox proportional hazards modeling with backwards stepwise selection (p ≤ 0.05 for entry; p > 0.10 for removal) including all the aforementioned demographic and clinical SEER variables. Multivariable Cox regression was performed both for the overall cohorts and within each T-stage category. Propensity score matching between ‘‘control’’ (i.e., appendectomy) and ‘‘case’’ (i.e., colectomy) was attempted but abandoned owing to significance imbalance between the two groups, highlighting the marked and nonrandom differences in treatment approaches in this national subset.17,18 Thus, only multivariable regressions were utilized. p-Values ≤ 0.05 were considered statistically significant; all tests were two-sided. Analyses were carried out using SPSS version 29.0 (IBM Corp., Armonk, NY).

Results

Rates and Predictors of Colectomy

A total of 7549 patients with appendiceal adenocarcinoma who underwent appendectomy or colectomy were identified in the SEER database. Serial exclusion of patients with metastatic disease at time of initial resection (n = 2843) and those without known disease-specific survival (n = 32) yielded a final cohort of 4674 patients. The majority (n = 3169; 67.8%) underwent colectomy; the remaining patients underwent appendectomy (n = 1505; 32.2%). Regarding histologic subtype, 2684 (57.4%) were NMAAs, and 1990 (42.6%) were MAAs.

In both the non-mucinous and mucinous subsets, significant clinicopathologic differences existed between the appendectomy and colectomy-treated groups (Table 1). For both histologies, colectomy was increasingly used for younger patients and was associated with higher T-stages. Notably, no disparities were observed by treatment strategy according to sex, median income, population density, and tumor grade.

Table 1 Clinicodemographics of NMAA and MAA patients in the overall study cohort
Full size table

Data on the number of lymph nodes examined were available in 4579 patients (98% of study cohort). Patients who underwent appendectomy had fewer lymph nodes examined (median 1, IQR 0–13) compared with colectomy patients (median 16, IQR 11–22; p < 0.001), including the subset of patients with N-positive (N+) disease (p < 0.001). In total, 161 patients underwent appendectomy alone for N-positive disease, and compared with colectomy patients (of any nodal staging) they were not significantly different with regards to age (p = 0.506), race/ethnicity (p = 0.555), median income (p = 0.974), or population density (p = 0.774). However, they were more likely to be male (61.5% versus 51.8%, p = 0.016) and more likely to receive adjuvant chemotherapy (64.6% versus 40.0%, p < 0.001).

Preoperative carcinoembryonic antigen (CEA) values were available for 25.6% of the overall cohort (non-mucinous: n = 670; mucinous: n = 525); in both subsets, elevated CEA was not associated with rates of colectomy (p = 0.400 and p = 0.605, respectively). In a multivariable logistic regression model, age, T-stage, and N-stage were independently associated with colectomy use (Supplemental Table 2).

Predictors of Lymph Node Positivity

Given that T-stage was associated with extent of surgery, the risk of lymph node metastases was quantified for increasing T-categories among colectomy patients. For non-mucinous histologies, the rate of nodal positivity increased with increasing T-stage (T1: 6.8%, T2: 11.4%, T3: 25.6%, and T4: 43.8%, p < 0.001). Likewise, for mucinous histologies, the rate of nodal positivity increased with increasing T-stage (T1: 4.6%, T2: 4.0%, T3: 17.1%, and T4: 21.6%, p < 0.001). Overall, the rate of nodal positivity was greater for non-mucinous histology (27.6% versus 16.4%, p < 0.001).

Since tumor grade has been previously reported to be associated with risk of lymph node disease, rates of lymph node metastases in NMAA and MAA were analyzed in our cohort, stratified by tumor grade.11,19 For both histologies, grade was significantly associated with N+ disease (p < 0.001), with poorly differentiated tumors exhibiting the highest rates of lymph node metastasis (NMAA: 44.4%, MAA: 31.3%, p = 0.002; Supplemental Table 3). Patients with well-differentiated and moderately differentiated tumors had lower rates of lymph node metastases, with no significant difference between NMAA and MAA tumors (well differentiated: 9.0% versus 7.7%, p = 0.450; poorly differentiated: 20.5% versus 19.2%, p = 0.490, respectively).

Influence of Colectomy on Disease-Specific Survival in the Overall Cohort

Median follow-up was 74 (IQR 33–131) months and 87 (IQR 41–150) months for the non-mucinous and mucinous adenocarcinoma subsets, respectively. A total of 1214 (26.0%) disease-specific deaths were observed. The long-term survival impact of extent of surgical resection was examined using Cox proportional hazards modeling, including those variables significantly associated with DSS by univariate analysis (Table 2). In both histologic subsets, variables independently associated with DSS included age, T-stage, N-stage, and grade. Utilization of colectomy was not associated with DSS in the non-mucinous [ hazard ratio (HR) 0.88, 95% confidence interval (CI) 0.75–1.03; adjusted p-value 0.107] nor mucinous (HR 0.95, 95% CI 0.78–1.17; adjusted p-value 0.963) subsets. Regardless of extent of surgical resection, prognosis was excellent for both non-mucinous [appendectomy: mean 164.1 (95% CI 155.4–172.9) months; colectomy: 168.4 (95% CI 162.6–174.3) months; log rank p = 0.209] and mucinous [mean 181.6 (95% CI 172.4–190.8) months; 181.7 (95% CI 175.3–188.1) months, log rank p = 0.801] adenocarcinoma patients.

Table 2 Cox regression model for independent predictors of disease-specific survival
Full size table

Survival Analysis of Extent of Resection, Stratified by T-stage

Given that T-stage and N-stage were each associated with disease-specific survival, and that nodal positivity was increasingly observed for higher T-stage, the impact of extent of surgery was analyzed for each T-stage. For non-mucinous appendiceal adenocarcinoma, there was no difference in DSS observed for appendectomy versus colectomy for patients with T1 tumors (log rank p = 0.298; Fig. 1). In contrast, utilization of colectomy was associated with improved DSS for patients with T3 (log rank p = 0.018) tumors and a trend toward improved DSS was observed in patients with T2 tumors (log rank p = 0.095). Patients with T4 non-mucinous tumors managed with appendectomy or colectomy evidenced similar DSS (log rank p = 0.912). In a multivariable cox regression accounting for age, N-stage, grade, and use of adjuvant chemotherapy, colectomy in patients with T3 tumors was associated with reduced risk of disease-specific death (HR 0.75, 95% CI 0.58–0.98, p = 0.032). Extent of surgery was not independently associated with DSS for other T categories (p = 0.495, p = 0.100, and p = 0.259 for T1, T2, and T4 lesions, respectively).

Fig. 1
figure 1

Impact of appendectomy versus colectomy on disease-specific survival in patients with NMAA, stratified by T-stage. NMAA non-mucinous appendiceal adenocarcinoma

Full size image

In contrast, for patients with mucinous tumors (Fig. 2), no difference in DSS was observed for appendectomy versus colectomy for any of the T-categories (T1: p = 0.568; T2: p = 0.952; T3: p = 0.825; and T4: p = 0.316). Multivariable Cox regression, accounting for age, N-stage, grade, and use of adjuvant chemotherapy, also demonstrated no difference between appendectomy and colectomy-treated groups (p = 0.169, p = 0.620, p = 0.761, and p = 0.310 for T1, T2, T3, and T4 lesions, respectively).

Fig. 2
figure 2

Impact of appendectomy versus colectomy on disease-specific survival in patients with MAA, stratified by T-stage. MAA mucinous appendiceal adenocarcinoma

Full size image

Survival Analysis of Extent of Resection, Stratified by Grade

The association of the extent of surgical resection and DSS was subsequently analyzed for each tumor type and stratified by tumor grade. For patients with moderately differentiated NMAA, hemicolectomy was associated with improved DSS compared with appendectomy (log rank p = 0.006; Fig. 3). In contrast, colectomy was not associated with DSS in well-differentiated (log rank p = 0.992) and poorly differentiated (log rank p = 0.762) NMAA. For patients with MAA, extent of resection was not associated with DSS for any tumor grade (well differentiated: p = 0.948, moderately differentiated: p = 0.289, poorly differentiated: p = 0.744).

Fig. 3
figure 3

Impact of appendectomy versus colectomy on disease-specific survival in patients with NMAA and MAA, stratified by grade. NMAA non-mucinous appendiceal adenocarcinoma, MAA mucinous appendiceal adenocarcinoma

Full size image

Discussion

The current study comprises the largest population-based study evaluating disease-specific survival by extent of resection for patients with mucinous and non-mucinous subtypes of appendiceal adenocarcinoma. In our analysis, non-mucinous histologies were associated with significant rates of lymph node metastases (26–44% for T3–4 lesions). In support of expert guidelines, surgical clearance of such lymph nodes with colectomy was associated with improvements in cancer-related deaths, particularly for T3 and moderately differentiated tumors. In contrast, mucinous histologies had lower rates of lymph node metastases, and use of colectomy was not associated with improved disease-specific survival for any MAA patients. Such data confirm the divergent biologic characteristics of mucinous versus non-mucinous AA and highlight the importance of consideration of histologic subtype in their surgical management.

Previous SEER-based propensity-score-matched analysis found that mucinous histology was not independently associated with long-term survival in stage I–III AA patients.20 The authors subsequently conclude that the same treatment strategies can be applied regardless of histologic subtype. However, similar risk-adjusted prognosis does not necessarily equate to similar surgical treatment. Here, by segregating the NMAA and MAA patients in all statistical analyses, a histology-dependent association between extent of resection and disease-specific survival was identified. For patients with NMAA, absolute improvements in DSS were observed for T2 and T3 lesions as well as moderately differentiated tumors. A DSS advantage was not observed in patients with T1 tumors. In support, previous studies have not demonstrated a survival benefit of an extensive lymphadenectomy for low risk T1 tumors, likely given the low rates of nodal disease.14,15 Patients with T1 tumors exhibited lymph node positivity rates of only 6.8% when undergoing colectomy. The lack of association between extent of surgery and survival for NMAA T4 tumors was an interesting finding, possibly driven by the higher risk of metastatic failure that would render extensive lymphadenectomy alone an ineffective way to achieve cure.

Oncologic tenets from the surgical management of colorectal cancer—namely, the use of colectomy for surgical staging of the draining lymph node basin—may not be relevant for the mucinous subtype of AA. In support, genomic analyses have revealed differences in the mutational landscape of AA and CRC as well as between MAA and NMAA.21 Compared with NMAA, MAA are more likely to harbor mutations in KRAS and GNAS, with fewer mutations in TP53.22,23 These genomic differences provide biologic support for consideration of distinct clinical entities requiring individualized treatment approaches. In these data, disease-specific survival was independent of type of resection for all MAA patients in a multivariable model adjusting for T-stage, grade, and use of adjuvant chemotherapy. Likewise, a previous analysis of MAA using the SEER dataset found that extent of resection was not associated with disease-specific survival, although this prior analysis involved a more heterogeneous cohort inclusive of patients with metastatic disease and data dating primarily from the twentieth century.24 Still, the current results were surprising, given the significant risk of nodal involvement for certain patients (17–22% in T3 and T4 MAA; 18–29% for moderately and poorly differentiated MAA).

One possible explanation could be the unique biologic behavior of mucinous tumors. These tumors have a propensity to recur within the peritoneum, which may not be prevented by more extensive lymphadenectomy. Mucin 2 (MUC2) is the most abundant, gel-forming, mucus protein primarily secreted in the small bowel and colon.25 MUC2 and MUC5A have been found to be overexpressed in MAA, with MUC2 being more profoundly expressed in pseudomyxoma peritonei of appendiceal origin.26,27 Overexpression of MUC2 has been associated with decreased survival in multiple cancer types.28,29,30,31 Under normal conditions, secreted mucin serves to protect the intestinal epithelium and subsequently undergoes degradation. However, when peritoneal seeding occurs, the produced mucin fails to degrade within the peritoneal cavity, leading to accumulation and development of pseudomyxoma peritonei. This mucin can shield cancer cells from the host’s immune system as well as prevent the delivery of chemotherapy agents, leading to treatment failure.32,33 Additionally, the mucin can facilitate the spread of tumor cells within the peritoneal cavity and create a favorable microenvironment that enhances tumor growth.33,34 This unique biologic behavior potentially renders an extensive lymphadenectomy alone an ineffective way to achieve cure for localized MAA. Relatedly, in a study examining outcomes in patients with metastatic MAA and peritoneal seeding who underwent cytoreductive surgery and intraperitoneal chemotherapy, the addition of right hemicolectomy did not provide a survival advantage over those who underwent appendectomy alone. Additionally, when the cohort was stratified by lymph node status, no difference in survival was observed between node-negative patients, node-positive patients, and patients with unknown nodal status.35

Without the prospect of randomized data given the rarity of this entity, these multivariable-adjusted registry-based data may be the best quality data available to guide clinical management. Still, these findings need to be interpreted with caution, and several limitations warrant emphasis. Cox regression models were used to adjust for known patient-related (i.e., race/ethnicity) and tumor-related confounders.8,22,36,37,38,39 Nevertheless, significant differences were observed between the appendectomy and the colectomy cohorts. An attempt to build propensity-score-matched cohorts failed, which points to the likely impact of treatment bias present in these nonrandomized data. Still, SEER reporting of disease-specific (versus overall) survival allowed for more precise understanding of the impact of surgical extent on long-term outcomes while minimizing the bias of patient comorbidity on both surgical treatment and the rate of non-disease-related deaths. A second major limitation is the lack of data on resection margin status following appendectomy as well as information on specific histologic characteristics, such as lymphovascular invasion (LVI). These two variables are important determinants in surgical management of T1 tumors; without these data, our findings may only be applicable for T1 tumors without positive resection margins or LVI. Third, the analyses of tumor grade are limited by institutional differences in pathologic grading as well as changes in the grade classification (i.e., two- versus three- versus four-tier system) that occurred over the study period. Thus, the results of the grade-stratified analyses need to be validated in additional contemporary series. Fourth, serum biomarkers (e.g., Ca 19-9 and CA 125) may be elevated in patients with appendiceal adenocarcinoma and have been associated with survival. Unfortunately, such data were not available to be included in these analyses. Lastly, limitations particular to SEER include certain database inadequacies, such as (a) specific chemotherapy regimens utilized, (b) completion of prescribed treatment schedules, (c) rates and types of recurrences, and (d) all clinical details that may have informed decisions regarding extent of surgery. In this regard, we observed a small number of patients who underwent appendectomy alone for node-positive disease that could not be explained by differences in their social determinants of health. It is possible that patient comorbidity may have contributed to some of these treatment decisions.

Conclusions

In this contemporary, population-based analysis of the use of colectomy for nonmetastatic mucinous and non-mucinous appendiceal adenocarcinoma, the survival impact of surgical strategies varied by histologic subtype. No improvement in disease-free survival was observed for the subset of patients with mucinous AA regardless of T-stage and histologic grade. These data support consideration of histologic subtype in the surgical treatment of localized AA. In the absence of randomized data, these results raise doubts regarding the need for colectomy for localized MAA.