Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy to Treat Pseudomyxoma Peritonei of Ovarian Origin: A Retrospective French RENAPE Group Study

Background Ovarian pseudomyxoma peritonei (OPMP) are rare, without well-defined therapeutic guidelines. We aimed to evaluate cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) to treat OPMP. Methods Patients from the French National Network for Rare Peritoneal Tumors (RENAPE) database with proven OPMP treated by CRS/HIPEC and with histologically normal appendix and digestive endoscopy were retrospectively included. Clinical and follow-up data were collected. Histopathological and immunohistochemical features were reviewed. Results Fifteen patients with a median age of 56 years were included. The median Peritoneal Cancer Index was 16. Following CRS, the completeness of cytoreduction (CC) score was CC-0 for 9/15 (60%) patients, CC-1 for 5/15 (33.3%) patients, and CC-2 for 1/15 (6.7%) patients. The median tumor size was 22.5 cm. After pathological review and immunohistochemical studies, tumors were classified as Group 1 (mucinous ovarian epithelial neoplasms) in 3/15 (20%) patients; Group 2 (mucinous neoplasm in ovarian teratoma) in 4/15 (26.7%) patients; Group 3 (mucinous neoplasm probably arising in ovarian teratoma) in 5/15 (33.3%) patients; and Group 4 (non-specific group) in 3/15 (20%) patients. Peritoneal lesions were OPMP pM1a/acellular, pM1b/grade 1 (hypocellular) and pM1b/grade 3 (signet-ring cells) in 13/15 (86.7%), 1/15 (6.7%) and 1/15 (6.7%) patients, respectively. Disease-free survival analysis showed a difference (p = 0.0463) between OPMP with teratoma/likely-teratoma origin (groups 2 and 3; 100% at 1, 5, and 10 years), and other groups (groups 1 and 4; 100%, 66.6%, and 50% at 1, 5, and 10 years, respectively). Conclusion These results suggested that a primary therapeutic strategy using complete CRS/HIPEC for patients with OPMP led to favorable long-term outcomes. Supplementary Information The online version contains supplementary material available at 10.1245/s10434-023-14850-0.

normal appendix and digestive endoscopy were retrospectively included.Clinical and follow-up data were collected.Histopathological and immunohistochemical features were reviewed.Results.Fifteen patients with a median age of 56 years were included.The median Peritoneal Cancer Index was 16.Following CRS, the completeness of cytoreduction (CC) score was CC-0 for 9/15 (60%) patients, CC-1 for 5/15 (33.3%) patients, and CC-2 for 1/15 (6.7%) patients.The median tumor size was 22.5 cm.After pathological review and immunohistochemical studies, tumors were classified as Group 1 (mucinous ovarian epithelial neoplasms) in 3/15 (20%) patients; Group 2 (mucinous neoplasm in ovarian teratoma) in 4/15 (26.7%) patients; Group 3 (mucinous neoplasm probably arising in ovarian teratoma) in 5/15 (33.3%) patients; and Group 4 (non-specific group) in 3/15 (20%) patients.Peritoneal lesions were OPMP pM1a/acellular, pM1b/grade 1 (hypocellular) and pM1b/grade 3 (signetring cells) in 13/15 (86.7%), 1/15 (6.7%) and 1/15 (6.7%) patients, respectively.Disease-free survival analysis showed a difference (p = 0.0463) between OPMP with teratoma/ Pseudomyxoma peritonei (PMP) is a clinical entity defined by the intraperitoneal accumulation of mucinous ascites and mucinous tumor deposits originating from an intra-abdominal primary tumor.2][3][4][5][6] This remains difficult to prove, as ovaries are commonly invaded in appendiceal PMP.[9][10] Accurate diagnosis of OPMP requires clinical, radiological and gastrointestinal endoscopic assessments as well as a histopathological examination of the appendix to exclude a primary digestive origin or an appendiceal mucinous neoplasia. 11A definitive diagnosis can only be performed after complete CRS, which includes the resection of the ovaries. 7hus, the absence of digestive tract tumor, of primary appendiceal neoplasia, and the identification of a mature teratoma or Brenner tumor associated with the mucinous tumor in the ovary, favor an ovarian origin of the PMP.According to the National Comprehensive Cancer Network (NCCN) guidelines, the current preferred regimens to treat ovarian mucinous carcinoma is 5-fluorouracil/leucovorin/oxaliplatin, or capecitabine/oxaliplatin, or paclitaxel/carboplatin for stage IC, with possible addition of bevacizumab in stages II-IV. 12][4][5][6]12 Although cytoreductive surgery (CRS) followed by hyperthermic intraperitoneal chemotherapy (HIPEC) is the standard treatment for PMP of appendiceal origin (APMP), [13][14][15] there are no current guidelines for the treatment of OPMP.][18][19][20][21][22][23][24][25][26][27] Recently, Yan et al. reported a case-series of eight OPMP; six of them were treated using CRS± HIPEC and there were no recurrences in two cases, which suggested a potential therapeutic benefit of this procedure. 10e objective of this study was to evaluate the efficacy of CRS/HIPEC to treat OPMP patients through a retrospective analysis of the French National Network for Rare Peritoneal Tumors (RENAPE) database, based on a comprehensive pathological review.The second objective was to assess whether the origin of the ovarian tumor cell (i.e., teratomatous, non-teratomatous-mucinous epithelium, or unknown origin) could have an impact on the efficacy of this treatment.

Selection Criteria and Study Design
Clinically suspected OPMP cases were retrospectively selected, from 2000 to 2023, using medical records recorded in the French RENAPE database, 28 labeled by the French National Cancer institute (INCa).According to the 2016 Peritoneal Surface Oncology Group International (PSOGI) classification consensus, OPMP were defined as an ovarian tumor with intraperitoneal accumulation of mucus including cellular or acellular mucinous ascites and peritoneal implants. 29All selected patients had normal colonoscopy and gastroscopy, an appendectomy, and underwent CRS/ HIPEC.Moreover, when magnetic resonance imaging (MRI) and/or computed tomography (CT) scans were available at diagnosis, the absence/presence of pancreato-biliary primary tumor or urachal tumor was also noted.
Patients with a PMP of non-ovarian origin after pathological review (i.e., APMP pathologically proven) or for which the appendicular origin could not be excluded (i.e., appendix not assessable by histopathology at the time of the study) were excluded.This was crucial to avoid confusion between OPMP and APMP. 14Patients without proven histopathological peritoneal mucin deposit were also excluded, as it was impossible to grade the pseudomyxoma.Similarly, patients with non-mucinous primary ovarian tumors (e.g., endometrioid carcinoma) were also excluded. 7

Clinical Data Collection
Clinical and therapeutic data were collected.The Peritoneal Cancer Index (PCI) and the completeness of cytoreduction (CC) score were rated, as previously defined. 15,30,31 Te quality of cytoreduction was defined according to the CC score: CC-0: no macroscopic residual tumor; CC-1: residual tumor <2.5 mm; CC-2: residual tumor between 2.5 mm and 25 mm; and CC-3: residual tumor >25 mm. 31 Cytoreductive surgery was considered complete when a CC-0 or CC-1 score was achieved.Intraperitoneal treatment consisted of HIPEC performed either with an open or closed abdomen technique, as previously described. 32ytoreductive Surgery and Hyperthermic Intraperitoneal … The postoperative morbidity and mortality were evaluated according to the Clavien-Dindo classification during the 90 postoperative days and according to the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE) version 5. [32][33][34] The intensive care unit (ICU) and total hospital length of stay were also recorded.Patients were followed-up every 3-6 months for 5 years, and then annually with clinical examination, serum blood tests and thoraco-abdomino-pelvic CT scan and/or peritoneal MRI completed with thoracic CT scan.Recurrence, defined as outcomes during follow-up, was diagnosed based on clinical, radiological, and/or histopathological findings, and was confirmed in multidisciplinary team meetings.The status at last news, total duration of follow-up, disease-free survival (DFS), and mean DFS were calculated.
All patients included in the French RENAPE database (NCT02834169) gave their informed consent.This study was conducted according to the Declaration of Helsinki.

Pathological Examination
All samples from patients with supposed OPMP, including primary ovarian tumor, appendix, and peritoneal samples, were retrospectively reviewed by several gynecopathologists (AT, MDS, CG) and expert digestive pathologists (NB, SI, JF). 35he following gross criteria were collected for each tumor: laterality, size of the primary mucinous ovarian tumor(s), and the presence of a capsule rupture.
Since two accepted classifications of tumors responsible for PMP are available (digestive system tumors and female genital tract tumors), 7,36 primary tumors were classified according to both: the 5th digestive system tumor WHO classification. 36as low-grade (grade 1) and highgrade (grade 2) mucinous neoplasms (LGMN and HGMN, respectively) or signet-ring cell adenocarcinoma (grade 3), and the 5th female genital tract tumor WHO classification. 7s mucinous cystadenoma, mucinous borderline tumor (± intraepithelial carcinoma, ± microinvasion, ± microcarcinoma), or invasive mucinous adenocarcinoma (expansive or infiltrative invasion pattern).The presence of the following histopathological criteria were also noted on the primary ovarian tumor: mucinous lesion continuum (from benign cystadenoma to borderline or malignant mucinous neoplasm), mature teratoma or a Brenner tumor component, areas of pseudomyxoma ovarii and glandular rupture and 'dirty' tumor necrosis in mucinous glands.
In all cases, the appendix was entirely included in paraffin blocks and histologically examined to exclude any appendiceal mucinous neoplasm.

Histopathological, Immunohistochemical and Clinical Correlation
Merging clinical, histopathological and immunohistochemical data, mucinous neoplasms responsive for OPMP were then classified into four distinct groups: Group 1: Arising from an ovarian mucinous epithelial origin (no teratoma component, ovarian-like immunohistochemical profile, and/or mucinous neoplasm associated with a Brenner tumor component).
Group 2: Arising from an ovarian teratoma (teratoma component with digestive-like immunohistochemical profile when available).
Group 3: Likely arising from an ovarian teratoma, in which teratoma was not sampled during gross examination or was overgrowth of the mucinous tumor (no teratoma component, digestive-like immunohistochemical profile, clinical data and imaging supporting a primary ovarian origin).
Group 4: Non-specific group (immunohistochemical data not available at the time of the study, or no specific profile).

Statistical Analysis
Descriptive statistics were used for the analysis of clinical and histopathological data, described as counts and percentages for dichotomous variables and medians and ranges for continuous values.
DFS was defined as the time from the first CRS/HIPEC procedure until recurrence or the last follow-up.Median time of follow-up and median DFS were calculated.
Survival analyses (DFS) were tested using the Kaplan-Meier method and log-rank test.The R statistical software (The R Foundation, Indianapolis, IN, USA; available from: http:// www.r-proje ct.org/) with the RcmdrPlugin.EZR package was used.A p-value <0.05 was considered statistically significant.The DFS calculation was performed to compare the survival between OPMP from teratoma/suspected teratoma origin (Groups 2 and 3) and ovarian/nonspecific origin (Groups 1 and 4), in order to assess whether the ovarian tumor cell origin could have an impact on the efficacy of this treatment.

Study Population
Between 2000 and 2023, 15 patients matching the inclusion criteria for OPMP were included (Fig. 1).The median age was 56 years (range 37.8-73.6)and the median preoperative CA125 dosage was 38 U/mL (range 22-800).At diagnosis, MRI and/or CT scans were available for 12/15 (80%) patients and showed no pancreato-biliary primitive or urachal tumor, while imaging performed during follow-up showed no pancreato-biliary primitive tumor.Two patients (13.3%) underwent a second surgical procedure (cases #13 and #14).At the time of the first CRS/HIPEC procedure, the American Society of Anesthesiologists (ASA) score was 1 in 3/14 (21.4%) patients, 2 in 8/14 (57.1%) patients, and 3 in 3/14 (21.4%) patients.The main characteristics of the patients and their treatment are fully reported in Table S1 of the electronic supplementary material (ESM) and summarized in Table 1.

Gross Examination of Primary Ovarian Tumors
A total of 13/15 (86.7%) patients had a unilateral mucinous tumor, while 2/15 (13.3%) had bilateral mucinous tumors that were associated with at least a teratoma component.For patients with unilateral tumors, the neoplasm was located in the left ovary in 5/13 (38.5%) patients, the right ovary in 6/13 (46.1%) patients, and not specified in 2/13 (15.4%) patients.
The median size of ovarian mucinous tumors was 22.5 cm (range 10-35), with all tumors ≥10 cm in diameter.All data are presented in ESM Table S2 and Table 1.

Histopathological Analysis Primary Ovarian Tumors
A mucinous tumor arising from a mature teratoma was found in 4/15 (26.7%) patients, while no component of mature teratoma was observed in 11/15 (73.3%) patients.However, a contralateral mature teratoma (not associated FIG. 1 Case selection process.All patients included in the present study had been treated by CRS followed by HIPEC.In order to ensure with certainty the primary ovarian origin, all patients included had to have a normal digestive endoscopic and imaging assessment and a complete histopathological examination of the appendix (which had to be available for analysis at the time of the study).In addition, OPMP had to be histologically confirmed with the presence of mucus with or without tumor cells in the peritoneal samples.The primary ovarian tumor had to be mucinous (epithelial or developed on a teratoma).* Colonoscopy and gastroscopy.CRS cytoreductive surgery, HIPEC hyperthermic intraperitoneal chemotherapy, OPMP Pseudomyxoma peritonei from ovarian origin, RENAPE French National Reference Center for Rare Peritoneal Tumors Cytoreductive Surgery and Hyperthermic Intraperitoneal … with a mucinous neoplasm) was found in 1/15 (6.7%) patients.An ipsilateral Brenner tumor was associated with the mucinous neoplasm in 1/15 (6.7%) patients.
Classifications of ovarian tumors according to the 5th WHO classifications of both genital female tumors. 7and digestive system tumors, 36 as well as histopathological analysis, are fully detailed in ESM Table S2   Peritoneal Samples PMP grading according to the 5th WHO classification of digestive system tumors. 36is presented in ESM Table S2 and Table 1.One patient (#5) had granulomatous peritonitis associated with PMP.

Histopathological Examination of the Appendix
No mucinous neoplasm of the appendix was observed after histopathological examination (data not shown).
The median length of hospital stay (conventional and ICU), median time of follow-up, and median DFS are presented in ESM Table S1 and Table 1.

DISCUSSION
PMP of ovarian origin is an extremely rare and difficult entity to diagnose.[9][10] The present study reported the largest series of patients with proven OPMP; all patients were successfully treated by CRS followed by HIPEC.A statistical difference was found between teratoma/suspected teratoma origin (groups 2 and 3) and ovarian/non-specific origin (groups 1 and 4) regarding the DFS.
To consider a PMP to be of ovarian origin, the cases were strictly selected; they had to present a normal digestive endoscopic assessment and a normal appendix after complete histopathological examination.A minority of the OPMP originated from an ovarian epithelial mucinous neoplasm with an ovarian-like expression profile (CK7>CK20; Group 1); one tumor was associated with a benign Brenner tumor.The majority were of probable teratomatous origin with a digestive-like expression profile (CK20>CK7; Groups 2 and 3), and a histologically visible mature teratoma was associated in four cases.In the remaining cases, the expression profile was not specific, and neither teratoma nor Brenner tumor was found.Thus, an extensive sampling and histopathological examination of the ovarian mucinous tumor is crucial, as the presence of a teratomatous component or a Brenner tumor would favor a PMP of ovarian origin. 38However, the mucinous malignancy can overgrow the initial mature teratoma; in one-third of cases, a digestivelike expression profile was found, without any teratomatous component.In the absence of primary digestive, appendicular, or pancreato-biliary tumor detected using endoscopy or imaging, as well as histologic appendicular examination, we speculated that these tumors might represent a mucinous neoplasia overgrowth in a mature teratoma.Moreover, the association of PMP with a granulomatous peritonitis around Cytoreductive Surgery and Hyperthermic Intraperitoneal … (f) HES (×100).On peritoneum, the pseudomyxoma peritonei consists of mucus dissociating the peritoneal tissue, accompanied by numerous siderophages (f corresponds to the black frame of e).(g) CK7 (×200).h CK20 (×200).(i) (SATB2; ×200).(j) PAX8 (×200).the mucinous neoplasm expressed CK7 intensely and diffusely, with a weak/focal expression of CK20, but no expression of PAX8 and SATB2 (ovarian-like profile).OPMP Pseudomyxoma peritonei from ovarian origin, HES hematoxylin-eosinsaffron keratin debris (Fig. 3) could also be considered as an argument for a teratomatous origin.In old cases, whose material was no longer available, immunohistochemistry could not be performed, while an IHC panel is systematically performed today.This panel should incorporate anti-CK7, anti-CK20, anti-PAX8, and anti-SATB2 antibodies. 39However, in Group 2 (teratomas), SATB2 was not positive in all tumors and PAX8 expression was focal/weak in one case.Herein, the expression of PAX8 did not enable the tumor classification, likely due to the small size of our series, suggesting that the most important immunohistochemical feature is the comparison of the CK7/CK20 expression.The criteria used for the classification of mucinous tumors are different in gynecological (ovarian mucinous (l) PAX8 (×200).The mucinous neoplasm expressed CK20 intensely and diffusely, with no expression of CK7 or PAX8 (digestive-like profile).OPMP Pseudomyxoma peritonei from ovarian origin, HES hematoxylin-eosin-saffron tumors) and digestive pathology (appendicular mucinous tumors), which complexify the discussion between pathologists, gynecologists, and surgeons.Although it was suggested to classify and grade mucinous tumors arising from ovarian teratomas as appendicular mucinous tumors, 40 it is not well defined as to how to classify and grade primary ovarian epithelial mucinous tumors responsible for OPMP (Group 1 herein).In the present study, tumors were classified according to the 5th edition of the WHO classification of both female genital tumors and digestive system tumors, 7,36 highlighting that LGMN were equivalent to mucinous cystadenomas and mucinous borderline tumors of the ovary, and that HGMN were equivalent to ovarian mucinous carcinomas (infiltrative pattern).Therefore, in the case of proven OPMP, it seems necessary to use both classifications, to allow all healthcare professionals (f) CK20 (×200).(g) SATB2 (×200).(h) PAX8 (×200).The CK20 expression was intense and diffuse in the mucinous neoplasm, with no expression of CK7 or PAX8.SATB2 expression was very focal and mild and was considered negative (digestive-like profile).OPMP Pseudomyxoma peritonei from ovarian origin, HES hematoxylin-eosin-saffron involved in the complex management of these patients to understand each other.However, no recommendations are currently available regarding the treatment of OPMP, as the disease is rare.In advanced ovarian cancer, platinum-based systemic chemotherapy is the standard regimen, regardless of the histological type.2][43] and a concordant poor prognosis, worse than in patients with nonresectable high-grade serous carcinoma treated with only platinum-based chemotherapy (overall survival [OS] of 14.6 vs. 40.8months; progression-free survival of 7.6 vs. 16.1 months), 44 reinforcing the importance of surgical and locoregional treatments.In the present series, the efficacy of the treatment using CRS/HIPEC on patients with OPMP was reported.Moreover, a better DFS was found for OPMP of teratoma/likely-teratoma origin compared with ovarian origin/non-specific origin (p = 0.0463).The difference in DFS between these groups may be due to the type of epithelial proliferation.Mucinous adenocarcinomas arising from a teratoma are of colonic/appendiceal epitheliumtype, while primary mucinous ovarian neoplasia are not supposed to arise from a colonic/appendiceal-type epithelium, as this does not exist in the ovary.4][15] Thus, it is not surprising that a better response to this regimen is found for OPMP of teratomatous origin (colonic/appendiceal-type epithelium) compared with OPMP of ovarian epithelial-type origin.In the literature, survival data for OPMP are scarce.Lee et al. reported a 5-year OS of 87% in a retrospective cohort of 35 cases, 45 most of whom were treated by CRS alone.However, appendicular status was not available as the materials used originated from ovarian lesions, not from the appendix.Yan et al. reported eight OPMP cases treated by CRS ± HIPEC, and only two patients were disease-free after 25 and 83 months of follow-up. 10Baratti et al. reported 19 cases of extra-appendicular PMP, among whom nine cases were of ovarian origin. 6; all patients underwent CRS/HIPEC and the 10-year OS for APMP and non-appendiceal origin was 63.4% and 62.0 %, respectively.Rufián-Andujar et al. reported 117 cases of PMP managed by CRS/HIPEC, of whom seven cases were of ovarian origin; for the OPMP subgroup, the reported 5-year OS was 80% and 5-year DFS was 66.7%. 26Herein, HIPEC regimens were heterogenous, and the majority were oxaliplatin and mitomycin alone.However, due to the high rates of severe morbidity, HIPEC protocols were replaced by cisplatin-based regimens, with a concomitant reduced risk of renal failure thanks to sodium thiosulfate intravenous perfusion. 15,46 ecently, the PSOGI led an international evidence-based consensus to define the recommended HIPEC protocol in such clinical situations (submitted).This study has limitations inherent to its retrospective design, such as the presence of missing data or materials.Another limitation is that the main patients reported herein presented OPMP with acellular peritoneal mucus deposits (pM1a), which likely contributed to an increased survival.Moreover, as OPMP is a rare disease, this case series was small in size and no control group could be identified within the French RENAPE group.Due to the limited data available on OPMP, mainly described as case reports, we were unable to find in the literature a series of OPMP treated by CRS alone to compare the results of the present series.One study reported two patients with OPMP treated with CRS alone (as patients could not receive HIPEC), followed by adjuvant chemotherapy and bevacizumab; one patient relapsed at 6 months, 21 highlighting the importance of a locoregional approach.In other case reports, surgery-alone treatment did not lead to relapses, but the duration of follow-up was short, 16,18,19 except for one patient followed-up 21 months after surgery. 23The addition of HIPEC could be especially useful in case of grade 3 PMP, which may relapse after surgical treatment alone. 24However, other studies are needed to confirm these hypotheses and to evaluate the oncological outcomes related to CRS/HIPEC, especially in multicenter studies, due to the rarity of this entity.

CONCLUSION
A diagnosis of OPMP should be established following a rigorous methodology in order to exclude other primary origins, namely appendiceal and pancreato-biliary.Considering the poor survival of patients with advanced mucinous ovarian tumors treated by systemic chemotherapy alone, a locoregional approach combining CRS+HIPEC should be considered as it was associated with long-term median DFS in this series.

FIG. 2
FIG. 2 Pathological and immunohistochemical features of OPMP developed on ovarian mucinous epithelial neoplasm (Group 1; case #2).(a) HES (×20).Mucinous tumor with complex architecture at low magnification and several components: (1) a first component of infiltrative mucinous adenocarcinoma with small glands composed of atypical cells infiltrating the stroma (b) HES (×200] corresponding to the red frame of (a): (2) a contingent of benign Brenner tumor consisting of transitional nest within a fibromatous stroma.(c) HES (×200) corresponds to the yellow frame of (a); and (3) a component of benign mucinous cystadenoma consisting of unistratified mucinous

FIG. 4
FIG. 4 Pathological and immunohistochemical features of OPMP developed on ovarian mucinous neoplasms, likely corresponding to ovarian teratoma not sampled during gross examination (Group 3; case #9).(a) Gross examination of ovarian tumor after formalin fixation: the multi-cystic tumor, poorly limited with external vegetations and mucus within the cysts.(b) HES (×10).The tumor consists of numerous cysts and glandular ruptures in the ovarian stroma.(c) HES (×80).The tumor consists of papillae lined with mucinous epithe-

FIG. 5
FIG.5 Survival analysis: comparison of the DFS of ovarian pseudomyxoma peritonei from teratomatous origin and other origins.The comparison of DFS between OPMP of teratoma/ suspected teratoma origin (Groups 2 and 3) and ovarian/ non-specific origin (Groups 1 and 4) showed a statistically significant difference (p = 0.0463).HIPEC hyperthermic intraperitoneal chemotherapy, DFS disease-free survival, OPMP Pseudomyxoma peritonei from ovarian origin and Table 1.Regarding the WHO genital female tumor classification, no expansive pattern was observed.