Timofeev, Julia MD; Galgano, Mary T. MD; Stoler, Mark H. MD; Lachance, Jason A. MD; Modesitt, Susan C. MD; Jazaeri, Amir A. MD
OBJECTIVE: To investigate the prevalence of appendiceal pathology in women undergoing surgery for a suspected ovarian neoplasm and the predictive value of intraoperative findings to determine the need for appendectomy at the time of surgery.
METHODS: Retrospective analysis of patients who underwent oophorectomy and appendectomy during the same surgical procedures at the University of Virginia Health System from 1992 to 2007. Observations were stratified based on the nature (benign, borderline, or malignant) and histology (serous compared with mucinous) of the ovarian neoplasm, frozen compared with final pathological diagnosis, and the gross appearance of the appendix.
RESULTS: Among the 191 patients identified, frozen section was consistent with seven mucinous and 35 serous carcinomas, 16 serous and 33 mucinous borderline tumors, 71 mucinous and serous cystadenomas, and 29 cases of suspected metastatic tumor from a gastrointestinal primary. The highest rates of coexisting appendiceal pathology were associated with serous ovarian cancers (94.4% of grossly abnormal and 35.3% of normal appendices) and ovarian tumors suspected to be of primary gastrointestinal origin (83.3% grossly abnormal and 60.0% normal appendices harbored coexisting mucinous neoplasms). Linear regression analysis revealed that appearance of the appendix and frozen section diagnosis of the ovarian pathology were statistically significant predictors of coexisting appendiceal pathology, but the latter was more important.
CONCLUSION: The prevalence of coexisting, clinically significant appendiceal pathology is low with a frozen section diagnosis of serous or mucinous cystadenoma. Appendectomy is recommended when frozen section diagnosis is mucinous or serous ovarian carcinoma, borderline tumor or metastatic carcinoma of suspected gastrointestinal origin.
LEVEL OF EVIDENCE: III
Coexistence of appendiceal and ovarian neoplasms is not rare and can result in diagnostic and treatment challenges. The appendix is a common site of metastasis from ovarian carcinoma, and appendectomy is frequently required to achieve optimal cytoreduction as part of ovarian cancer surgery. Appendectomy may also be performed prophylactically in the setting of benign or malignant ovarian neoplasms. Rationales for prophylactic appendectomy include concerns regarding possible occult metastatic ovarian cancer and prevention of acute appendicitis given its potential morbidity in patients undergoing chemotherapy. The role of routine prophylactic appendectomy as part of ovarian cancer staging is still controversial, particularly in early-stage disease, with several studies recommending against it as a result of low risk of appendiceal involvement.1–3 In contrast, other authors advocate routine removal of appendix as a result of the possibility of upstaging of early-stage ovarian cancer or simply as part of complete staging and maximal cytoreduction.4–7 Most studies agree that the rate of appendiceal metastasis is higher in patients with stage III or IV disease, which has been reported to be as high as 83%.1,2,4,6,8
Another consideration in performing appendectomy is to ensure accurate diagnosis in cases in which apparent ovarian mucinous carcinomas, in fact, represent metastatic disease arising from an occult appendiceal primary.9–12 Such cases may mimic early- or advanced-stage mucinous ovarian cancer with or without pseudomyxoma peritonei. Further complicating the diagnosis is the fact that histologic appearance of gastrointestinal tumor metastases to ovaries may be indistinguishable from a primary ovarian mucinous borderline (low malignant potential) tumor. These considerations have led some clinicians to advocate prophylactic appendectomy in the setting of any mucinous ovarian neoplasm.
The primary objective of this study was to determine whether information available at the time of surgery, namely the gross appearance of the appendix and the frozen section pathology impression, can be used to predict the coexistence of appendiceal pathology and, thus, the need for appendectomy. As a secondary aim, we estimated the prevalence of coexisting appendiceal pathology (based on the final surgical pathology report) with various subtypes of ovarian neoplasms.
MATERIALS AND METHODS
After receiving approval of the University of Virginia's Institutional Review Board for Health Sciences Research, the University of Virginia Clinical Data Repository database was searched to identify all female patients who have undergone an oophorectomy and an appendectomy during single surgery for pelvic mass from 1992 to 2007. Pathology reports were then reviewed for information pertaining to frozen pathology, gross involvement of the appendix (as described either in operative note or pathology report), and the final ovarian and appendiceal diagnoses. Ovarian frozen section diagnoses were subdivided into mucinous and serous carcinomas, tumors of low malignant potential, cystadenomas, and tumors suspected to represent metastasis from a gastrointestinal primary. Rates of coexisting appendiceal pathology were then calculated for each subgroup. Gross appendiceal abnormalities included thickening of the wall, hyperemia, adhesions, and metastatic implants. Pathologic findings in the appendix included carcinoid, mucocele, appendicitis, appendiceal low malignant potential tumors, metastatic implants, and appendiceal carcinoma. Rates of occult and gross appendiceal pathology were then calculated for each subgroup. Patients who did not have a gross description of the appendix intraoperatively, when no frozen section was done, or any patients without specific mention of the appendiceal diagnosis on final pathology were excluded. Tumors that were carcinomas or “at least” low malignant potential on frozen section were reviewed by two gynecologic pathologists to ensure uniform classification.
Statistical analysis was performed using DTREG Predictive Modeling software. The ability of the gross appearance of the appendix and ovarian frozen pathology diagnosis to predict appendiceal pathology was assessed using linear regression analysis with singular value decomposition. A P value of <.05 was considered statistically significant. The importance of predictor calculation was performed using sensitivity analysis in which the values of each variable are randomized and the effect on the quality of the model is measured. The most important variable is assigned an arbitrary value of 100 and other variables are assigned corresponding relative values.
To determine whether the intraoperative assessment of the pelvic mass (through frozen section pathology of ovarian neoplasm) and the appendix (based on gross appearance) was useful in guiding the decision to perform appendectomy, we reviewed 191 patients whose surgery included both salpingo-oophorectomy and appendectomy. The demographic and clinical characteristics of the subjects and the nature of the pelvic mass are presented in Table 1. Among these cases, the frozen section diagnosis revealed seven mucinous and 35 serous carcinomas, 16 serous and 33 “at least” mucinous low malignant potentials (with definitive diagnosis deferred to final pathology), 71 mucinous and serous cystadenomas, and 29 cases of suspected metastatic disease from a primary gastrointestinal cancer (Table 2).
Of the 71 patients with serous or mucinous cystadenomas, 59 had grossly normal appendices with pathology significant for a single incidental carcinoid (1.7%). Twelve of 71 had grossly abnormal-appearing appendices. Four of these 12 had pathologic findings on microscopic examination, which included one carcinoid, two cases of subclinical appendicitis, and one case of endosalpingiosis involving the appendix.
Among patients with a frozen section diagnosis of “at least” mucinous ovarian borderline tumor and a grossly normal appendix, the rate of occult appendiceal pathology was two of 31 (6.5%) consisting of two incidental carcinoids. Two patients in this group had grossly abnormal-appearing appendices but neither had any significant pathologic findings. Sixteen patients had intraoperative diagnosis of serous low malignant potential tumors. None of the eight grossly normal appendices had pathologic findings, whereas two of eight (25.0%) abnormal appendices had evidence of involvement by the serous low malignant potential tumor (one noninvasive macroscopic serosal implant and one Psammoma body).
Seven patients received the intraoperative diagnosis of mucinous ovarian carcinoma based on frozen section; the appendix appeared grossly abnormal in one case and final histology revealed ovarian cancer metastasis to the appendiceal serosa. The remaining six patients had no pathologic abnormalities in the appendix on final pathology. Thirty-five patients were diagnosed with serous ovarian carcinoma, 18 (51.4%) had a grossly abnormal appendix, and 17 of 18 (94.4%) were positive for metastatic carcinoma. Moreover, even when the appendix appeared grossly normal, 35.3% had microscopic evidence of ovarian cancer metastases on final pathology. It is noteworthy, however, that all cases of metastatic serous carcinoma involving the appendix occurred in patients with other evidence of stage III or IV disease.
Not surprisingly, patients with a frozen section diagnosis consistent with “suspected metastatic gastrointestinal malignancy” had higher rates of gross abnormalities of the appendix (24 of 29 [82.8%]) and were also likely to have occult disease (three of five [60.0%]) when the appendix appeared grossly normal (Table 1).
To determine if intraoperative findings, namely the gross appearance of the appendix and frozen section pathologic diagnosis of the ovarian neoplasm, could predict appendiceal pathology, a linear regression model consisting of these two variables was constructed. The strength of each variable for predicting the final appendiceal pathology is represented by its regression coefficient (β). The most clinically relevant parts of this analysis are presented in Table 3. Overall, in this model the ovarian frozen section diagnosis was much more important than the gross appearance of the appendix for predicting the coexistence of appendiceal pathology (relative scores of 100 compared with 23.3, respectively). This is mainly the result of the presence of occult appendiceal pathology in grossly normal appendices as well as the fact that many abnormal appendices represented fibrosis, appendicitis, adhesions, or other normal variations.
For prevalence estimations, only final pathologic diagnoses were considered. Seventy-six patients had mucinous ovarian neoplasms and included 14 patients with carcinomas, 26 low malignant potential tumors, and 36 cystadenomas. The prevalence of appendiceal pathology among patients with mucinous cystadenomas was one of 36 (2.7%). This patient had an incidental 2-mm carcinoid tumor on histologic evaluation. Among patients with mucinous low malignant potential tumors, only one of 26 (3.8%) had pathologic findings in the appendix and this was a 1-cm appendiceal carcinoid tumor. In patients with ovarian mucinous carcinomas, one of 14 (7.1%) had an appendiceal serosal metastasis from ovarian cancer.
Among 88 patients with serous ovarian neoplasms, there were 41 carcinomas, 12 serous low malignant potentials, and 35 serous cystadenomas. Rates of coexisting appendiceal pathology were: zero of 35 cystadenomas, one of 12 (8.3%) low malignant potentials, and 27 of 41 (65.9%) carcinomas. One patient in the serous low malignant potential tumor group had a tumor implant on the appendix in the setting of stage III disease. All 27 patients with positive appendiceal findings in the carcinoma group had metastatic papillary serous carcinoma on final pathology. Among patients with a diagnosis of mucinous ovarian carcinoma, only one of 14 (7.1%) had stage III or IV disease in contrast to serous ovarian carcinomas in which 38 of 41 (92.7%) of patients had late-stage disease on final pathology. There was no occult appendiceal carcinoma in early-stage cancers.
Twenty-seven patients had a final diagnosis consistent with primary gastrointestinal tumor. Of these, 22 of 27 (81.5%) had appendiceal involvement on final pathology.
This is the first investigation to specifically address the rates of coexisting appendiceal pathology based on information available at the time of surgical exploration for pelvic mass or suspected ovarian cancer. These data provide a rational basis for consideration of an appendectomy by the treating surgeon. In addition, our results add to and extend previous data on appendiceal involvement by metastatic ovarian cancer by specifically addressing this issue in cases of ovarian low malignant potential tumors and cystadenomas. We observed a significant rate of appendiceal metastases associated with serous ovarian carcinomas (65.9%) with all cases occurring in patients with other evidence of stage III or IV disease. In contrast, no appendiceal pathology was noted in patients with serous cystadenomas and only one of 12 (8.3%) serous borderline tumors was noted to have an implant of serous low malignant potential on the appendix.
Our study does have a number of limitations. Only the carcinomas and low malignant potential tumors were reviewed by gynecologic pathologists. However, it is unlikely that a review of benign neoplasms would have led to a significant change in their classification. Furthermore, our investigation only included patients who underwent both appendectomy and salpingo-oophorectomy during the same surgery; our results may be biased toward overestimating the rates of coexisting appendiceal pathology. Despite this limitation, our findings are in line with previously published reports. Malfetano9 reported an appendiceal metastasis rate of 70% among patients with stage III or IV epithelial ovarian cancer; however, no histologic stratification was provided. Fontanelli et al2 reported on 160 patients with 23% having metastasis to the appendix (all of them stage III and IV) with 91% of tumors in this study classified as serous cell type. Rose et al1 reviewed records on 80 patients who had an appendectomy during primary debulking surgery with rates of appendiceal metastasis ranging from 48% in serous and 8% in mucinous malignancies. Poorly differentiated and serous carcinomas involved the appendix in 69.7% of stage III and IV cases, and borderline tumors were noted to have metastasis to the appendix in 8% of patients.
A consistent finding in all these studies is that the vast majority of appendiceal metastases arise from stage III and IV serous ovarian cancers. In support of this conclusion, Ramirez et al11 found no occult or macroscopic disease of the appendix among 57 patients with early-stage ovarian cancer (81% had stage I and 19% had stage II disease). However, in this series, 40% had pure mucinous tumors, and 25% had ovarian tumors of low malignant potential. In contrast, Ayhan et al3 evaluated 285 patients with epithelial ovarian carcinoma and found 37% of patients had metastasis to the appendix. Furthermore, histologic findings of the appendix resulted in upstaging of 4.9% of patients, leading these authors to advocate routine appendectomy as part of cytoreductive surgery.
One of the objectives of our investigation was to determine if mucinous ovarian neoplasms are associated with higher prevalence of coexisting appendiceal pathology, possibly justifying routine appendectomy in these cases. In fact, we observed that among 62 patients with cystadenomas or borderline tumors of mucinous histology, there were only two cases of coexisting appendiceal pathology, both of which were incidental small carcinoid tumors. This low prevalence of coexisting appendiceal pathology was comparable to that observed with serous benign and borderline neoplasms. Overall, based on the final pathology diagnosis, there were no cases of coexisting appendiceal carcinoma and three cases of incidental carcinoid tumors (all 1 cm or less in size) among 109 patients with benign or borderline ovarian neoplasms. As noted, it is likely that the true prevalence of appendiceal pathology is even lower as a result of the patient selection bias in our study.
Finally, our study supports the notion that the decision to perform appendectomy can be made based on frozen section diagnosis and the gross appearance of the appendix, although the former factor is more important. We anticipate that any grossly abnormal appendix would be removed regardless of ovarian pathology. We found a low prevalence of clinically significant, coexisting appendiceal pathology in patients with serous or mucinous benign or borderline ovarian neoplasms. In contrast, appendiceal pathology is more frequently present when the frozen diagnosis is consistent with mucinous or serous carcinoma and when the appendix is grossly abnormal. When frozen pathology of the ovarian neoplasm indicates mucinous or serous ovarian carcinoma, low malignant potential tumor, or metastatic carcinoma of suspected gastrointestinal origin, appendectomy is recommended (even if the appendix is grossly normal). Among benign neoplasms (mucinous and serous), appendectomy is indicated only if gross abnormalities are present. It should be noted that despite the low rate of significant appendiceal pathology associated with a diagnosis of mucinous low malignant potential ovarian tumor on frozen section (Table 2) and the fact that this variable was not significantly predictive for the coexistence of a mucinous appendiceal neoplasm (Table 3), we still recommend routine appendectomy in these cases. This is based on the relatively small number of low malignant potential mucinous samples included in our study as well as the diagnostic challenges associated with frozen section evaluation of these frequently large ovarian tumors.
1. Rose PG, Reale FR, Fisher A, Hunter RE. Appendectomy in primary and secondary staging operations for ovarian malignancy. Obstet Gynecol 1991;77:116–8.
2. Fontanelli R, Paladini D, Raspagliesi F, di Re E. The role of appendectomy in surgical procedures for ovarian cancer. Gynecol Oncol 1992;46:42–4.
3. Ayhan A, Gultekin M, Taskiran C, Salman MC, Celik NY, Yuce K, et al. Routine appendectomy in epithelial ovarian carcinoma: is it necessary? Obstet Gynecol 2005;105:719–24.
4. Sonnendecker EW. Is appendicectomy mandatory in patients with ovarian carcinoma? S Afr Med J 1982;62:978–80.
5. Hristov AC, Young RH, Vang R, Yemelyanova AV, Seidman JD, Ronnett BM. Ovarian metastases of appendiceal tumors with goblet cell carcinoidlike and signet ring cell patterns: a report of 30 cases. Am J Surg Pathol 2007;31:1502–11.
6. Dietrich CS 3rd, Desimone CP, Modesitt SC, Depriest PD, Ueland FR, Pavlik EJ, et al. Primary appendiceal cancer: gynecologic manifestations and treatment options. Gynecol Oncol 2007;104:602–6.
7. Gehrig PA, Boggess JF, Ollila DW, Groben PA, Van Le L. Appendix cancer mimicking ovarian cancer. Int J Gynecol Cancer 2002;12:768–72.
8. Ronnett BM, Kurman RJ, Shmookler BM, Sugarbaker PH, Young RH. The morphologic spectrum of ovarian metastases of appendiceal adenocarcinomas: a clinicopathologic and immunohistochemical analysis of tumors often misinterpreted as primary ovarian tumors or metastatic tumors from other gastrointestinal sites. Am J Surg Pathol 1997;21:1144–55.
9. Malfetano JH. The appendix and its metastatic potential in epithelial ovarian cancer. Obstet Gynecol 1987;69:396–8.
10. Bese T, Kosebay D, Kaleli S, Oz AU, Demirkiran F, Gezer A. Appendectomy in the surgical staging of ovarian carcinoma. Int J Gynaecol Obstet 1996;53:249–52.
11. Ramirez PT, Slomovitz BM, McQuinn L, Levenback C, Coleman RL. Role of appendectomy at the time of primary surgery in patients with early-stage ovarian cancer. Gynecol Oncol 2006;103:888–90.
12. Westermann C, Mann WJ, Chumas J, Rochelson B, Stone ML. Routine appendectomy in extensive gynecologic operations. Surg Gynecol Obstet 1986;162:307–12.
Figure. No caption available.