Mucinous Low Malignant Potential Research Articles

Mucinous ovarian cancer: A rare entity with variable clinical presentation and management (148)

Mucinous ovarian cancer: A rare entity with variable clinical presentation and management (148)

Abstract B25: Hormone receptor expression and response in ovarian tumors of low malignant potential

Abstract Objective: Primary treatment for ovarian low malignant potential (LMP) tumors is surgical resection with little proven effect of cytotoxic chemotherapy. New therapeutic strategies are needed to prevent and treat recurrence. The objective of this study is to investigate the role of hormone receptors in ovarian low malignant potential tumors. Methods: We obtained formalin-fixed, paraffin embedded (FFPE) tissue samples from patients with ovarian LMP tumors. Clinicopathologic data were collected for all available specimens. Immunohistochemistry (IHC) was performed to determine protein expression. Gene expression was measured using the NanoString nCounter platform. Differences in expression between histologic subtypes were by t-test, followed by a Bonferroni adjustment for multiple comparisons (p<0.01). Results: Thirty-eight patients were included in this study. The histologic subtypes were divided between 20 serous and 18 mucinous tumors. All serous tumors were ERα and PR positive by IHC and 6 of 20 (30%) were ERβ positive. Only one of the 18 mucinous tumors (5.6%) was positive for both ERα and PR, and all the samples were negative for ERβ. ERβ mRNA levels were elevated in the mucinous LMP tumors and ERα mRNA was higher in the serous samples. ERα and PR protein expression correlated with mRNA expression of ESR1 (r2=0.5606) and PR (r2=0.42), respectively. ERβ expression did not correlate with mRNA expression of ESR2 (r2=0.038). Measurement of estrogen-regulated genes and genes involved in ER signaling (n=236) revealed a clear separation of serous from mucinous subtypes, with 44 genes showing a statistically significant difference (p<0.01) between the serous and mucinous subtypes. Conclusion: High levels of ERα, PR, and downstream ERα targets in serous LMP tumors suggest hormone responsiveness and that these tumors are potentially treatable with endocrine therapy. Mucinous tumors largely did not express hormone receptors. In light of this, differences in gene expression observed are more likely attributable to underlying lineage differences than differences in estrogen signaling. Additional large-scale analyses are warranted to better understand the differences between subtypes and to identify avenues for targeted therapy in mucinous LMPs. Citation Format: Sarah E. Taylor, Courtney L. Andersen, Rohit Bhargava, George Tseng, Steffi Oesterreich, Robert P. Edwards, Adrian V. Lee. Hormone receptor expression and response in ovarian tumors of low malignant potential. [abstract]. In: Proceedings of the AACR Conference: Addressing Critical Questions in Ovarian Cancer Research and Treatment; Oct 1-4, 2017; Pittsburgh, PA. Philadelphia (PA): AACR; Clin Cancer Res 2018;24(15_Suppl):Abstract nr B25.

Abstract 2998: MicroRNA expression signatures in human ovarian tumor progression

Abstract Differential expression of a number of miRNAs has previously been detected between normal ovaries and epithelial ovarian cancer (EOC). However, miRNA expression signatures have not been demonstrated between primary human ovarian surface epithelial (HOSE) cell, cystadenoma, low malignant potential (LMP) ovarian tumor and ovarian carcinoma as well as fallopian tube. Here, we performed miRNA profiling in normal primary OSE cells, normal fallopian tube and ovarian tumors, including 7 cystadnomas, 6 LMP and 49 ovarian carcinomas, as well as 15 ovarian cancer cell lines and 5 HIOSE cell lines. While deregulated dozen miRNAs were observed in different histological types of cysadenoma, LMP and ovarian carcinoma compared to primary OSE and fallopian tube, there are no significantly common changed miRNAs between serous adenoma and LMP, as well as between mucinous adenoma and LMP. However, 3 miRNAs were co-downregulated and 7 miRNAs were co-upregulated in serous LMP and carcinoma. When comparing mucinous LMP and carcinoma, only 3 miRNAs were co-deregulated. In addition, 3 and 6 miRNAs were co-deregulated in mucinous and serous adenoma and carcinoma, respectively. While miRNA expression patterns are distinct among four histological types of ovarian carcinoma examined, upregulation of miR-221/222 family was found to have significant poor prognosis value in serious carcinoma (p<0.05). These results support the notion that majority of ovarian carcinoma are derived neither from cystadenoma nor from LMP. Further cluster analysis suggests that the serious ovarian cancer may be derived from fallopian tube or/and OSE. Also our data suggest that commonly deregulated miRNAs between adenoma, LMP and carcinoma could serve as not only early markers but also initiation pathogenetic factors in this malignancy. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 2998.

Estrogen and progesterone receptor status and outcome in epithelial ovarian cancers and low malignant potential tumors

Objective Receptors for estrogen (ER) and progesterone (PR) are prognostic indicators for a variety of endocrine tumors including breast and endometrial. This study was conducted to determine if ER and PR expression patterns are predictive of outcome in patients with epithelial ovarian cancer (EOC) or ovarian low malignant potential (LMP) tumors. Methods ER and PR protein levels were assessed by immunohistochemistry in 45 LMP and 89 EOC samples. Patterns of ER/PR expression (individually and combinations of ER−/PR−, ER+/PR−, ER−/PR+, and ER+/PR+) were correlated with standard prognostic factors of overall survival (OS) in this patient population. Results For patients with EOC, the 5-year OS per ER−/PR+, ER+/PR−, ER+/PR+, and ER−/PR− expression was 83%, 79%, 61%, and 48%, respectively, and these differences were statistically significant. In multivariate analyses, ER/PR expression patterns were found to be independent predictors of OS, as were the classical prognostic factors of grade, stage, debulking, and chemotherapy response to treatment. In patients with mucinous LMP tumors, ER and PR were absent. Because no LMP patients died of disease during the studied period, no correlation analysis with OS could be performed. Conclusions Patterns of ER/PR expression provide prognostic information in EOC. Additional studies evaluating hormonal inhibition may help personalize the therapy of patients with ovarian cancer.

Validation of Candidate Serum Ovarian Cancer Biomarkers for Early Detection

ObjectiveWe have previously analyzed protein profiles using Surface Enhanced Laser Desorption and Ionization Time-Of-Flight Mass Spectroscopy (SELDI-TOF-MS) [Kozak et al. 2003, Proc. Natl. Acad. Sci. U.S.A. 100:12343–8] and identified 3 differentially expressed serum proteins for the diagnosis of ovarian cancer (OC) [Kozak et al. 2005, Proteomics, 5:4589–96], namely, apolipoprotein A-I (apoA-I), transthyretin (TTR) and transferin (TF). The objective of the present study is to determine the efficacy of the three OC biomarkers for the detection of early stage (ES) OC, in direct comparison to CA125.MethodsThe levels of CA125, apoA-I, TTR and TF were measured in 392 serum samples [82 women with normal ovaries (N), 24 women with benign ovarian tumors (B), 85 women with ovarian tumors of low malignant potential (LMP), 126 women with early stage ovarian cancer (ESOC), and 75 women with late stage ovarian cancer (LSOC)], obtained through the GOG and Cooperative Human Tissue Network. Following statistical analysis, multivariate regression models were built to evaluate the utility of the three OC markers in early detection.ResultsMultiple logistic regression models (MLRM) utilizing all biomarker values (CA125, TTR, TF and apoA-I) from all histological subtypes (serous, mucinous, and endometrioid adenocarcinoma) distinguished normal samples from LMP with 91% sensitivity (specificity 92%), and normal samples from ESOC with a sensitivity of 89% (specificity 92%). MLRM, utilizing values of all four markers from only the mucinous histological subtype showed that collectively, CA125, TTR, TF and apoA-I, were able to distinguish normal samples from mucinous LMP with 90% sensitivity, and further distinguished normal samples from early stage mucinous ovarian cancer with a sensitivity of 95%. In contrast, in serum samples from patients with mucinous tumors, CA125 alone was able to distinguish normal samples from LMP and early stage ovarian cancer with a sensitivity of only 46% and 47%, respectively. Furthermore, collectively, apoA-I, TTR and TF (excluding CA-125) distinguished i) normal samples from samples representing all histopathologic subtypes of LMP, with a sensitivity of 73%, ii) normal samples from ESOC with a sensitivity of 84% and iii) normal samples from LSOC with a sensitivity of 97%. More strikingly, the sensitivity in distinguishing normal versus mucinous ESOC, utilizing apoA-I, TF and TTR (CA-125 excluded), was 95% (specificity 86%; AUC 95%).ConclusionsThese results suggest that the biomarker panel consisting of apoA-I, TTR and TF may significantly improve early detection of OC.

Ovarian epithelial tumors of low malignant potential: Are they precursors of ovarian carcinoma?

Ovarian tumors of low malignant potential (LMP) are intermediate between benign adenoma and frank ovarian carcinoma, and are characterized by the absence of destructive stromal invasion, or microinvasion. If LMP did not develop into invasive carcinoma, then a minimally invasive simple treatment could be chosen. However, histological diagnosis cannot completely predict the prognosis of patients with ovarian tumors. It was found that mucinous LMP is most frequent in Japan, whereas serous LMP is the most common in Western countries. Mucinous LMP with loss of heterozygosity (LOH) at chromosomes 5q14-21 and 17q11.2, and serous LMP were suggested to be precursor lesions of ovarian carcinomas. Follow-up study revealed that patients with mucinous LMP who had LOH at 19q12 and/or Xq11-12 were at the greatest risk of progression. It is concluded that molecular genetic analysis such as LOH study could predict prognosis and aid in treatment decision-making. The present review describes molecular genetic changes of LMP and presents problems on LMP that remain to be elucidated.

Expression Profiling of Mucinous Tumors of the Ovary Identifies Genes of Clinicopathologic Importance

To elucidate the molecular mechanisms contributing to the unique clinicopathologic characteristics of mucinous ovarian carcinoma, global gene expression profiling of mucinous ovarian tumors was carried out. Gene expression profiling was completed for 25 microdissected mucinous tumors [6 cystadenomas, 10 low malignant potential (LMP) tumors, and 9 adenocarcinomas] using Affymetrix U133 Plus 2.0 oligonucleotide microarrays. Hierarchical clustering and binary tree prediction analysis were used to determine the relationships among mucinous specimens and a series of previously profiled microdissected serous tumors and normal ovarian surface epithelium. PathwayAssist software was used to identify putative signaling pathways involved in the development of mucinous LMP tumors and adenocarcinomas. Comparison of the gene profiles between mucinous tumors and normal ovarian epithelial cells identified 1,599, 2,916, and 1,765 differentially expressed in genes in the cystadenomas, LMP tumors, and adenocarcinomas, respectively. Hierarchical clustering showed that mucinous and serous LMP tumors are distinct. In addition, there was a close association of mucinous LMP tumors and adenocarcinomas with serous adenocarcinomas. Binary tree prediction revealed increased heterogeneity among mucinous tumors compared with their serous counterparts. Furthermore, the cystadenomas coexpressed a subset of genes that were differentially regulated in LMP and adenocarcinoma specimens compared with normal ovarian surface epithelium. PathwayAssist highlighted pathways with expression of genes involved in drug resistance in both LMP and adenocarcinoma samples. In addition, genes involved in cytoskeletal regulation were specifically up-regulated in the mucinous adenocarcinomas. These data provide a useful basis for understanding the molecular events leading to the development and progression of mucinous ovarian cancer.

Recurrent ovarian carcinoma of low malignant potential: The role of secondary surgical cytoreduction and the prognosis in Chinese patients

The study of pathology and prognosis for patients with ovarian tumors of low malignant potential (LMP) has been well documented. The treatment, particularly the secondary surgery, for recurrent disease is less documented. This study was conducted to investigate the role of secondary surgical cytoreduction for recurrent ovarian carcinoma of LMP. Patients with ovarian carcinoma of LMP at advanced stages experiencing disease recurrence and treated by secondary surgical cytoreduction at Fudan University Cancer Hospital were retrospectively reviewed. Sixteen patients with recurrent serous and mucinous ovarian LMP tumors who underwent secondary surgical cytoreduction were entered in this study. The median age was 46.5 years. The 5-year survival in those with residual disease < or =1 cm after secondary surgical cytoreduction was 83%, compared to 26% in those with residual disease >1 cm (P = 0.01). Multivariate analysis suggested that extent of recurrent disease (solitary vs. multiple, relative risk [RR] =2.69, P < 0.01), and residual disease after secondary surgical cytoreduction (< or =1 cm vs. >1 cm, RR = 2.56, P < 0.01) were determinants of survival. For recurrent ovarian carcinomas of low malignant potential, prolonged survival was observed in patients with solitary recurrent disease and optimal secondary surgical outcome.

Allelic loss at 19q12 and Xq11–12 predict an adverse clinical outcome in patients with mucinous ovarian tumours of low malignant potential

Ovarian tumours of low malignant potential (LMP) are intermediate between adenomas and ovarian carcinomas. These tumours are often associated with a significantly better prognosis than ovarian carcinomas. However, a subset of these tumours can progress and become lethal. In order to seek sensitive diagnostic tools for monitoring patients after surgical operation, we performed a genome-wide scan for loss of heterozygosity (LOH) in 41 mucinous LMPs using 91 polymorphic microsatellite markers at an average interval of 50 cM across all of the human chromosomes and 25 LOH markers reportedly associated with ovarian carcinoma. In addition, we assessed whether clinicopathological parameters, microvessel density, Ki-67 labeling index, apoptotic index or p53 overexpression would be useful for predicting the postoperative outcome of LMP patients. Of the 116 markers examined, 19q12 and Xq11–12 showed significant correlation between postoperative progression-free survival time and LOH status (P<0.05). Patients with a high Ki-67 labeling index had a significantly poorer progression-free survival time than those with lower levels (P=0.042). Other clinicopathological factors and immunohistochemical analysis had no correlation with progression-free survival time in this series of patients. When the combination of LOH at 19q12 and/or Xq11–12 was assessed using Cox's regression analysis, patients with tumours that showed LOH at these positions were at greatest risk of progression (P=0.0073). These findings suggest that the identification of LOH at 19q12 and/or Xq11–12 in former mucinous LMP sites should alert the clinician to the presence of a potentially aggressive lesion in the coelomic epithelium, even if a distinction between second primary tumours or recurrence could not be determined.

Increased MCL-1 expression is associated with poor prognosis in ovarian carcinomas.

To investigate the potential role of the BCL–2 gene family (BAX, BCL–2, MCL–1, and BCL‐XL) in ovarian cancer development and progression, mRNA expression levels of these genes were measured using semi‐quantitative PCR in epithelial ovarian tumor tissues and normal ovaries. The immunohistochemical expression of MCL–1 in ovarian tumors was also examined. The expression levels of BAX and MCL–1 mRNA were significantly higher in ovarian cancers and in adenomas than in normal ovaries (P<0.05). In contrast, the BCL–2 mRNA expression level in ovarian cancers was significantly lower than in ovarian adenomas and in normal ovaries (P<0.05). Expression of BCL‐XL mRNA was no different between normal ovaries and ovarian tumors. Log‐rank testing showed that low BAX mRNA expression and high MCL–1 mRNA expression significantly correlate with poor survival for patients with stage III ovarian carcinomas (BAX, P=0.05; MCL–1, P=0.02). Immunohistochemical analysis showed that diffuse‐positive expression of MCL–1 protein in mucinous carcinomas was significantly higher than in mucinous low malignant potential (LMP) tumors (P=0.03). In ovarian cancer cases, diffuse‐positive expression of MCL–1 protein significantly correlates with advanced clinical stage, high histologic grade, and poor survival (stage, P<0.01; grade, P=0.01; survival, P=0.01). These results suggest that increased MCL–1 expression may play an important role in replacing the functions of increased BAX and decreased BCL–2 in ovarian carcinoma cells, thereby promoting cell survival, and resulting in a poor prognosis for patients with ovarian cancer.

Comprehensive allelotype study of ovarian tumors of low malignant potential: potential differences in pathways between tumors with and without genetic predisposition to invasive carcinoma.

Ovarian tumors of low malignant potential (LMP) are intermediate between adenomas and ovarian carcinomas (OC); however, the relevance of LMP to ovarian carcinogenesis is not clear. We performed a comparative analysis of allelotypes in 50 cases of LMP (42 mucinous and 8 serous) and 23 cases of OC (15 mucinous and 8 serous) to investigate any differences in genetic changes. Analysis of loss of heterozygosity (LOH) using 25 microsatellite markers reportedly associated with OC revealed that the total LOH frequency at each marker was significantly lower in LMP than in OC (p < 0.01). However, 9 (36%) loci showed higher LOH frequency in mucinous LMP than in mucinous OC. A genome-wide scan for LOH using 91 microsatellite markers and fine mapping revealed that LOH at D7S1805 (7q35) is characteristic of mucinous LMP (19.4% in mucinous LMP, 8.3% in mucinous OC). We further studied LOH in 3 cases of mucinous OC that were accompanied by mucinous LMP lesions. In 2 cases, LOH frequency was higher in the carcinoma portion than in the morphologically LMP portion. The other case showed microsatellite instability in the morphologically LMP portion and LOH in the carcinoma portion. Our results suggest the presence of an LMP-to-OC developmental sequence and the existence of a subset of LMP that does not develop into OC in the mucinous subtype of ovarian tumors.

Numerical abnormalities of chromosomes 1, 11, 17, and X are associated with stromal invasion in serous and mucinous epithelial ovarian tumours.

The clinical behaviour of ovarian tumours of low malignant potential (LMP) is unpredictable and it has been suggested that the majority of these lesions have no invasive potential. This study has analysed 92 epithelial ovarian tumours [11 mucinous cystadenomas, 18 mucinous LMP tumours, 15 mucinous carcinomas (9 FIGO stage I), 16 serous cystadenomas, 15 serous LMP tumours, and 17 serous carcinomas (11 FIGO stage I)] for numerical abnormalities of chromosomes 1, 11, 17, and X by interphase cytogenetics. Overall, numerical aberrations were identified in none of the cystadenomas, 15 per cent of serous LMP tumours, 17 per cent of mucinous LMP tumours, 67 per cent of mucinous carcinomas, and 82 per cent of invasive serous carcinomas. In mucinous LMP tumours, chromosome gains were associated with spindled nuclear morphology. Chromosome abnormalities were significantly more frequent in invasive mucinous (overall p< 0.01; stage I p< 0.05) and serous (overall p< 0.001; stage I p< 0.01) carcinomas than in the corresponding LMP tumours. No significant relationship between either stromal invasion or tumour type and the pattern of chromosome loss or gain was identified, although monosomy X was identified almost exclusively in invasive serous carcinomas. These observations are consistent with the concept that LMP tumours are unlikely to be precursors of ovarian carcinoma, but suggest that chromosome instability is important in the development of the invasive phenotype.

Peritoneal Washing Cytology of Ovarian Tumors of Low Malignant Potential

To correlate the cytologic diagnoses of peritoneal washings (PWs) with histologic findings of ovarian tumors of low malignant potential (LMP). PW cytology from 90 patients with histologically confirmed ovarian tumors of LMP were reviewed and correlated with histologic findings. In 90 LMP tumor patients who underwent PW, cytologic diagnoses were reported as positive in 30 (33%) cases, atypical in 5 (6%) cases and negative in 55 (61%) cases. On review of the five atypical cases, three were reclassified as negative and two as positive. Peritoneal washings were positive in 20 (33%) of 60 patients with serous LMP tumors, 12 (44%) of 27 patients with mucinous LMP tumors and none of 3 patients with mixed LMP tumors. The presence of positive cytology was highly indicative of surface ovarian involvement or peritoneal implants in completely staged or selectively biopsied patients (P < .0001). Of 60 (67%) patients who had a staging laparotomy, positive PWs were seen in 3 (75%) of 4 cases with surface involvement only, 13 (94%) of 14 cases with peritoneal implants only, 3 (100%) of 3 cases with both peritoneal implants and ovarian surface involvement, and 9 (23%) of 39 cases without evidence of surface involvement or peritoneal implants. PWs were positive in 4 (13%) of 30 cases without staging laparotomy. PW cytology is a sensitive indicator of peritoneal dissemination of ovarian tumors of LMP. In addition, PWs will detect a high percentage of patients with subclinical intraperitoneal extension of such tumors and should be used routinely in the staging of ovarian tumors of LMP.

Rare mutations and no hypermethylation at the CDKN2A locus in epithelial ovarian tumours.

The tumour-suppressor gene CDKN2A (p16, MTS1, CDK4I) encodes a cell cycle-regulatory protein and is located on chromosome 9p21, a region deleted in a wide variety of human cancers. To determine the role of the CDKN2A gene in the development of ovarian adenocarcinomas, we examined a large series of benign, low malignant potential (LMP) and invasive ovarian neoplasms for evidence of loss of heterozygosity (LOH), homozygous deletions, point mutations and hypermethylation of the CDKN2A locus. We have previously reported LOH on 9p in 45% of malignant ovarian neoplasms and a smaller percentage of benign and LMP tumours. In the current study, 6 malignant tumours were identified with partial deletions of 9p21. In 5 of these, the CDKN2A gene lays within the minimal deleted region. Homozygous deletions of CDKN2A were observed in only 2/88 invasive ovarian tumours and in 5/11 ovarian cancer cell lines. Of 15 primary ovarian tumours analyzed, one nonsense mutation was identified in a mucinous LMP tumour. No evidence of hypermethylation of the CDKN2A gene was found in 50 primary ovarian adenocarcinomas nor in 3 ovarian cancer cell lines. In conclusion, homozygous deletions, mutations and the de novo methylation of 5' CpG island are not frequent modes of inactivation of the CDKN2A gene in ovarian cancer. The target of 9p LOH in ovarian adenocarcinomas is therefore unknown.

Microinvasion in low malignant potential tumors of the ovary

The presence of microinvasion (Mi) has not previously been investigated in nonserous low malignant potential (LMP) tumors of the ovary. In serous LMP tumors (SLMP), Mi has not worsened the prognosis compared with usual SLMP in previous reports. In a retrospective clinicopathologic review of 126 cases of serous and mucinous LMP (MLMP) tumors of the ovary, the authors identified 14 cases with Mi: seven of 72 SLMP, four of 44 mucinous intestinal LMP (MILMP), and three of 10 mucinous müllerian LMP (MMLMP). Tumors with Mi resembled usual LMP histologically except for small foci (up to 0.2 cm) in the stroma consisting predominantly of single cells or small clusters of cells, sometimes in a cribriform pattern. In SLMP-Mi, the cells had eosinophific cytoplasm, and the stromal reaction was minimal/absent. In MILMP-Mi and MMLMP-Mi, nuclear staining was paler; the stroma often showed a fibrous reaction or edema, especially in MILMP-Mi. All 14 patients had follow-up (2.2 to 18.6 years) and were without evidence of disease. Careful screening of both mucinous and serous ovarian LMP can show Mi. Mi foci should be distinguished from true stromal invasion, pseudoinvasion, endothelial cells, decidual cells and histiocytes, and pseudomyxoma in mucinous LMP. Mi does not seem to worsen prognosis in MLMP and SLMP tumors of the ovary.

Pseudomyxoma peritonei in women: A clinicopathologic analysis of 30 cases with emphasis on site of origin, prognosis, and relationship to ovarian mucinous tumors of low malignant potential

Pseudomyxoma peritonei (PMP) is a poorly understood condition characterized by the accumulation of abundant mucinous material within the peritoneal cavity and associated with a mucinous tumor of the gastrointestinal tract or ovaries. Recently there has been considerable debate over the primary site of origin of the tumor associated with PMP in women. Some investigators have proposed a primary site in the ovaries, whereas others favor the gastrointestinal tract or the peritoneum. Another confusing issue has been the nature of the ovarian mucinous tumors associated with PMP. Although these neoplasms may be frankly malignant, more often they show minimal cytologic atypia and epithelial proliferation and have been classified as borderline or low malignant potential tumors. In order to address the issues of site of origin and nature of the associated ovarian mucinous tumors, we studied 68 cases of PMP in women, 30 of whom had mucinous tumors involving the ovaries. All 30 of these cases had an associated mucinous appendiceal or intestinal tumor. The PMP cases with ovarian tumors were compared with 30 ovarian mucinous tumors of low malignant potential (LMP). Based on the analysis of the primary ovarian mucinous LMP tumors, a set of criteria was formulated and used to determine the probable site of origin of PMP in the 30 women with mucinous tumors involving the ovaries. The following gross and microscopic features of the ovarian tumor were considered to be inconsistent with a primary ovarian origin: (1) surface involvement with or without superficial stromal involvement only; (2) adenocarcinoma with signet ring cell differentiation, with a previously diagnosed or concurrent appendiceal tumor of similar morphology; (3) bilateral adenocarcinoma consistent with colonic or appendiceal morphology; and (4) unilateral adenocarcinoma consistent with colonic or appendiceal morphology with a history of a co-. Ionic or appendiceal adenocarcinoma. When any one of these features was present the ovarian tumor was diagnosed as secondary. The following additional features also were considered to be more typical of secondary ovarian involvement: (1) normal or only slightly enlarged ovaries; (2) bilateral ovarian involvement; (3) simple or only focally proliferative mucinous epithelium with abundant extracellular mucin in cases with predominantly surface involvement of the ovaries, with or without a history of/or concurrent appendiceal adenoma; (4) multifocal or extensive pseudomyxoma ovarii in cases with stromal involvement, with or without a history of/or concurrent appendiceal adenoma; (5) ruptured appendiceal adenoma and unruptured ovarian tumor of similar histology; and (6) presence of an associated mucinous intestinal tumor. If at least three of the above features were present, the ovarian tumor also was diagnosed as secondary. When classified according to all of these criteria 28 of the 30 cases of PMP (93.3%) were attributed to primary mucinous appendiceal or colonic tumors, and the remaining two were compatible with either intestinal or ovarian origin. No cases were of unequivocal ovarian origin. Analysis of multiple histologic features of the ovarian, intestinal, and peritoneal mucinous tumors showed that the PMP cases could be separated into two pathologically and prognostically distinct groups based on the associated intestinal tumor diagnosis. Appendiceal adenomaassociated cases showed a significantly better prognosis than appendiceal or intestinal carcinoma-associated cases ( P < .0001). Nearly all appendiceal adenoma-associated cases were characterized by appendiceal, ovarian, and peritoneal tumors with no significant atypia or epithelial proliferation and absence of lymph node involvement and parenchymal organ invasion. In contrast, the carcinoma-associated cases displayed significant atypia and epithelial proliferation in the intestinal, ovarian, and peritoneal tumors. In addition, these tumors were frequently associated with lymph node and parenchymal organ metastases. Accordingly, we recommend that the term PMP be restricted to those cases that have the features of appendiceal adenoma-associated cases. The carcinoma-associated cases should be classified as mucinous carcinomas and not included in the category of PMP. Finally, the conclusion that PMP is virtually always of appendiceal origin provides the basis for reassessing the entire concept of ovarian mucinous LMP tumors because malignant behavior associated with the latter tumors has virtually always been confined to women with PMP.