Survival outcomes of appendiceal mucinous neoplasms by histological type and stage: Analysis of 266 cases in a multicenter collaborative retrospective clinical study

Abstract Aim Appendiceal mucinous neoplasms are rare, and thus the literature is sparse with regard to histological types, staging, and prognosis. In particular, it is unclear how long‐term outcome may differ between mucinous adenocarcinomas and other adenocarcinomas. In the present study, we aimed to investigate the histological types and stages of appendiceal neoplasms, and to evaluate the prognostic impacts of these factors in patients with mucinous adenocarcinomas and non‐mucinous adenocarcinomas. Methods Patients with appendiceal tumors diagnosed between 2007 and 2016 were retrospectively identified from the databases of 19 institutions in the Clinical Study Group of Osaka University, Colorectal Group. Results A total of 266 patients with appendiceal tumors were identified, of whom 130 had pathologically diagnosed adenocarcinomas, including 57 with mucinous adenocarcinomas and 73 with non‐mucinous adenocarcinomas. Five‐year overall survival (OS) rates were 64.5% for mucinous adenocarcinomas, and 49.0% for non‐mucinous adenocarcinomas. OS was significantly shorter among patients with non‐mucinous adenocarcinomas compared to mucinous adenocarcinomas. Among patients with mucinous adenocarcinomas, 5‐year OS rates were 53.6% for stage 0/I, 82.6% for II/III, and 48.4% for IV. Among patients with non‐mucinous adenocarcinomas, 5‐year OS rates were 90.9% for stage 0/I, 68.8% for II/III, and 7.1% for IV. Analysis of patients with stage IV disease revealed significantly shorter OS among patients with non‐mucinous adenocarcinomas compared to mucinous adenocarcinomas. Conclusion Our present findings showed a better prognosis in patients with mucinous adenocarcinomas compared to non‐mucinous adenocarcinomas. In this setting, Union for International Cancer Control staging was associated with prognosis for non‐mucinous adenocarcinomas, but not for mucinous adenocarcinomas.


| INTRODUC TI ON
The vermiform appendix is the primary site of several distinctive benign and malignant neoplasms. Appendiceal tumors are rare neoplasms, comprising approximately 1% of appendectomy specimens. 1 The major categories of primary appendiceal neoplasms include epithelial tumors (subclassified as mucinous tumors, neuroendocrine tumors, and mixed glandular and endocrine tumors 2 ), mesenchymal tumors, and lymphomas. Mucinous adenocarcinomas are the most common non-carcinoid tumors of the appendix. 3 Mucinous neoplasms of the appendix constitute a heterogeneous group of neoplasms, ranging from adenomas to mucinous adenocarcinomas. 4 Various systems for classifying appendiceal mucinous neoplasms have been proposed by various authors [5][6][7][8] and in the World Health Organization (WHO) 2010 guidelines. 9 In an effort to simplify the diagnostic terminology for appendiceal mucinous neoplasms, the WHO has identified morphologic characteristics that can be used to classify low-grade and high-grade tumors. 9 Although our understanding of appendiceal mucinous neoplasms has advanced, their classification remains confusing.
A recent review provides an updated clarification of the various classification systems. 10,11 The Peritoneal Surface Oncology Group International (PSOGI) has published their consensus regarding the classification and proposed diagnostic terminology for primary appendiceal mucinous neoplasms, which provides rigorous diagnostic criteria for low-grade appendiceal mucinous neoplasms (LAMNs). 10 They recommend using the 2016 Modified Delphi Consensus Protocol to classify non-carcinoid epithelial appendiceal tumors into eight histomorphological architectural groups: adenoma, serrated polyp, LAMN, high-grade appendiceal mucinous neoplasm (HAMN), mucinous adenocarcinoma (well/moderately/poorly differentiated), signet ring cell low-differentiated (mucinous) adenocarcinoma, signet-ring cell (mucinous) adenocarcinoma, and adenocarcinoma.
On the other hand, The American Joint Committee on Cancer (AJCC) 8th edition clarifies LAMN staging to include prognostically relevant criteria, and describes a new T category specifically for LAMN, termed Tis(LAMN). 12 Both the AJCC 8th edition and the PSOGI consensus emphasize the importance of distinguishing between low-grade and high-grade intraperitoneal disease, and both advocate for three-tier grade assessment of appendiceal mucinous neoplasms, in which low-grade tumors are classified as G1, while high-grade tumors are classified as G2 or G3. 12 Appendiceal adenocarcinoma is defined by the presence of infiltrative invasion, 10 and can be subdivided into mucinous, non-mucinous, and signet-ring cell histological types. 13 Appendiceal adenocarcinoma seems to have different characteristics from other colorectal cancers, but sufficient evidence is lacking due to its rarity. However, the AJCC 8th edition now includes separate classifications for appendiceal carcinomas and colorectal carcinomas. 12 To date, little is known about variations in long-term outcomes, but it is expected that long-term postoperative performance differs between mucinous adenocarcinoma and other adenocarcinomas.
In the present multicenter retrospective clinical study of appendiceal tumors, we aimed to investigate the histological types and stages

Conclusion:
Our present findings showed a better prognosis in patients with mucinous adenocarcinomas compared to non-mucinous adenocarcinomas. In this setting, Union for International Cancer Control staging was associated with prognosis for non-mucinous adenocarcinomas, but not for mucinous adenocarcinomas.

K E Y W O R D S
appendiceal carcinoma, mucinous, non-mucinous, prognosis, survival outcomes of appendiceal neoplasms. We further evaluated the prognostic impacts of these factors in patients with mucinous adenocarcinoma and non-mucinous adenocarcinomas of the appendix.

| Study design
We retrospectively analyzed a cohort of patients who had been histologically diagnosed with appendiceal tumors between January 2007 and December 2016. This study was approved by our Institutional Review Board (approval number 17019).

| Data source and study population
Data for this study were acquired from the medical records of 19 institutions participating in a multicenter collaborative research group (Clinical Study Group of Osaka University, Colorectal Group).
Patients with unknown survival data, and those not coded as benign tumors (including LAMN) or adenocarcinoma, were retrospectively excluded from further analysis ( Figure 1).
All recorded clinical and pathological data were validated against medical and pathology records. Recorded variables included age, sex, surgery, surgical stages (one-stage/two-stage), surgical approach, surgical procedure, lymph node dissection, combined resection, histological grade, and TNM stage according to the Union for International Cancer Control (UICC) 8th version. 14 The types of surgical procedure included appendectomy, cecal resection, ileocecal resection (ICR), right hemicolectomy, others, and unknown. The types of operations included one-stage, two-stage, others, and unknown.

| Endpoints
The study endpoint was overall survival (OS), defined as the time in months from the date of surgery to the date of death from any cause.
Recurrent disease was diagnosed based on clinical and pathological findings, laboratory results, and diagnostic imaging.
F I G U R E 1 Study flow chart

| Statistical analysis
We performed descriptive data analyses, calculating frequencies and percentages for categorical variables, and expressing continuous variables as median (range). We evaluated the significance of betweengroup differences using the Mann-Whitney U and Kruskal-Wallis tests. The Kaplan-Meier method was used to determine the effects of each variable on survival, and log-rank tests were used to compare survival curves. Hazard ratios (HR) were reported as point estimates with a 95% confidence interval (CI). All statistical analyses were performed using JMP Pro Version 12 (SAS Institute, Inc., Cary, NC, USA).
The median age at diagnosis was higher among patients with nonmucinous adenocarcinomas compared to those with benign tumors and mucinous adenocarcinoma. Regarding operation type, Patients with benign tumors had higher rates of one-stage surgery, laparoscopic approach, appendectomy, and R0 resection than the other groups. The group with mucinous adenocarcinomas had a higher proportion of females and higher rate of combined resection than the other groups. Compared to patients with benign tumors and mucinous adenocarcinomas, patients with nonmucinous adenocarcinomas were older and had higher rates of emergency surgery and open approach.

| Survival
During follow-up, 39 patients (16.9%) died from appendiceal tumors, and there were a total of 49 (21.3%) all-cause deaths. In two cases, death occurred ≥70 months after diagnosis, even with benign tumors. Two patients diagnosed with LAMN had tumor-related deaths more than 5 years after surgery. In one of these cases, the patient underwent R2 resection, and was diagnosed with LAMN (T3N0M1b) with disseminated pseudomyxoma peritonei (PMP). In the other case, the patient underwent ICR with combined resection of ovary, uterus, and rectum, and was also diagnosed with LAMN   Figure 4). Regarding TNM categories, mucinous and non-mucinous adenocarcinomas did not significantly differ in the rates of pathological T stage (P = 0. 12) or N stage (P = 0.14).  Figure S1).

| D ISCUSS I ON
The results of our present multicenter retrospective clinical study highlight the interplay between histological types and stages among patients with appendiceal neoplasms, and show how these factors impact prognosis between mucinous and non-mucinous adenocarcinomas. In two cases, death occurred more than 5 years after surgery, even with benign tumors. These findings suggest that these cases should be considered mucinous adenocarcinoma due to the diagnosis of LAMN. Our results also revealed a better prognosis in patients with mucinous adenocarcinomas compared to non-mucinous adenocarcinomas in this setting. Furthermore, higher UICC stage was associated with increased risk of death among patients with non-mucinous adenocarcinomas, but not with mucinous adenocarcinomas.
There is presently scarce evidence regarding the appropriate management and surveillance of patients with appendiceal mucinous neoplasms. LAMN is by far the most frequent source of PMP, and can develop intraperitoneal recurrence many years after the initial presentation. 7,[15][16][17] Therefore, patients with LAMN should routinely be offered follow-up to monitor for subsequent PMP development. 18 In fact, our data showed that two patients with LAMN died due to their disease more than 5 years after their initial sur- Unfortunately, our present data cannot clarify the relationship between appendiceal mucinous adenocarcinomas and PMP. PMP is a clinical syndrome in which a mucinous neoplasm grows within the peritoneal cavity causing mucinous ascites and peritoneal implants. 19 It is uncommon, with an incidence of approximately 0.2 per 100 000 per year. 17 Actually, no recurrence cases due to PMP were revealed among appendiceal mucinous adenocarcinomas in this study. However, in the present setting we analyzed a cohort of patients who had been histologically diagnosed with appendiceal tumors, and no patients with carcinomas of unknown primary are included. Furthermore, the diagnostic criteria of PMP and appendiceal mucinous adenocarcinomas have not yet become sufficiently understood in general among pathologists, despite accurate diagnosis of both appendiceal mucinous adenocarcinomas and PMP being so important. 19 Thus, it is difficult to clarify the incidence rate of death due to PMP among appendiceal mucinous adenocarcinomas.
Consistent with previous reports, 13,20 we identified a number of differences between mucinous and non-mucinous adenocarcinomas.  Female sex and disease stage IV were more common for mucinous adenocarcinomas. On the other hand, stage III disease was less common for mucinous adenocarcinomas than non-mucinous adenocarcinomas, suggesting that mucinous adenocarcinomas showed preferential peritoneal spread rather than lymphatic spread. In a recent large retrospective study of mucinous and non-mucinous colorectal adenocarcinomas, mucinous adenocarcinomas were more commonly found at a more advanced stage, and were predominantly located at the right side of the colon. 21 They also reported that patients with mucinous adenocarcinomas were younger than those with non-mucinous adenocarcinomas, 21  The present study has both strengths and limitations. One of its strengths is that the dataset was based on the medical records from each institution participating in a multicenter collaborative research group with the aim of investigating the histological types and stages of appendiceal neoplasms. This enabled us to evaluate precise data relating to mucinous adenocarcinomas and non-mucinous adenocarcinomas, which may be difficult to identify in a larger cohort.
One limitation of our study is its retrospective design. It is unknown whether the study included patients with mixed-type adenomas.
Additionally, the number of patients was small and the follow-up period was not matured. Therefore, our data may not be sufficient to draw definitive conclusions, compared to previous studies with larger cohorts. [27][28][29] It should also be noted that patients with missing survival data were excluded from survival analysis, which could potentially introduce bias.
Another potential limitation of this study is the high percentage of mucinous adenocarcinomas with unknown differentiation, which was significantly higher than the percentage of non-mucinous adenocarcinomas with unknown differentiation.
Moreover, the 67.8% rate of unknown differentiation among mucinous adenocarcinomas was significantly higher than the rates of 32% and 35% reported in the National Cancer Data Base and SEER databases, respectively. [27][28][29] Thus, the evidence from our present study could not support the current AJCC 8th and PSOGI clas-

| CON CLUS ION
Here we found that mucinous adenocarcinomas are associated with better prognosis than non-mucinous adenocarcinomas. In the present setting, UICC staging was associated with prognosis among non-mucinous adenocarcinomas, but not mucinous adenocarcinoma. Analysis of stage IV disease revealed better prognosis in patients with mucinous adenocarcinoma compared to non-mucinous adenocarcinomas. Further studies are needed to confirm these findings.

ACK N OWLED G M ENTS
We would like to thank Yuriko Takeda, Supporting Center for Clinical Research and Education (SCCRE) Data Center, for their work on data management. This work was supported by the SCCRE.

D I SCLOS U R E
Conflict of interest: the authors have no conflicts of interest to declare.