When high-grade serous carcinoma involves the fallopian tube and either ovary or peritoneum, the ovary is typically considered the primary site, partly because of its much greater frequency as a primary site compared with fallopian tube primaries, yet studies aimed at identifying a precursor lesion within the ovary have been unsuccessful.2,23 Recently, it has been suggested that a subset of carcinomas that appear to be primary ovarian or peritoneal high-grade serous carcinoma may be of fallopian tube origin based on identification of high-grade intraepithelial serous carcinomas in the fallopian tube of women at high risk of developing ovarian cancer, as well as in women with sporadic (nonhereditary) ovarian carcinoma.7,12,14,17
To determine the validity of this hypothesis and address other issues relating to tubal origin of ovarian carcinoma, we studied the pathologic features of pelvic carcinomas with and without tubal intraepithelial carcinomas (TICs) in a consecutive series of in-house cases that qualified as primary ovarian, peritoneal, and tubal carcinomas based on conventional criteria.
MATERIALS AND METHODS
All consecutive in-house cases of pelvic (nonuterine) gynecologic carcinomas at The Johns Hopkins Hospital (Baltimore, MD) from 2006 through 2008 were reviewed (n=114). Fifty-two carcinomas of ovarian, peritoneal, and fallopian tube origin, in which all visible fallopian tube tissue was resected [and completely submitted for histologic examination per the protocol for sectioning and extensively examining the fimbriated end of the fallopian tube (SEE-FIM protocol18)], were identified and are the subject of this study.
Criteria for Assigning Primary Site of Origin
Tumors were classified as primary ovarian, peritoneal, or tubal after review of hematoxylin and eosin (H&E) slides, gross pathology reports, and clinical notes in The Johns Hopkins Hospital computer system based on conventional criteria as detailed below.
Carcinomas were classified as peritoneal in origin using Gynecologic Oncology Group criteria3 if: both ovaries were not enlarged by tumor; the extent of nonovarian disease was greater than that involving the ovaries; microscopically, the ovaries showed either no tumor or tumor in parenchyma/stroma <5 mm in greatest dimension; and the extraovarian tumor displayed a histologic appearance consistent with that seen in ovarian serous carcinomas.
Tumors were classified as ovarian in origin if none of the above criteria were fulfilled. In addition, tumors that were ≥5 mm on the ovarian surface (unless the tumor was imbedded in a desmoplastic plaque characteristic of secondary ovarian involvement) were classified as primary ovarian, even if no tumor was present in ovarian parenchyma/stroma.
Carcinomas were classified as primary tubal in origin if tumor arose from the endosalpinx and the size of the fallopian tube tumor was larger than the ovarian tumor.11,21 Other features that have been used previously, such as the histologic appearance of the tumor resembling tubal mucosa and transition from benign to malignant epithelium,11,21 were considered unreliable and, therefore, not used in this study as high-grade serous carcinomas do not resemble the bland ciliated mucosa of normal fallopian tube and because a transition from benign to malignant epithelium may not be apparent in all tubal primaries.
Assessment of Pathologic Features
TIC was defined as noninvasive tubal epithelium exhibiting marked nuclear atypia characterized by loss of polarity, increased nuclear size, increased nuclear-to-cytoplasmic ratios, hyperchromasia, and irregular nuclear membranes and chromatin distribution (Figs. 1–3); in addition, absence of cilia and mitotic figures also characterized TICs. Epithelial stratification and nuclear molding were useful but not present in all cases. Although criteria in the literature for the distinction of TIC from noninvasive intraluminal high-grade serous carcinoma are ill defined, the TICs in this study did not create an intraluminal mass. Pathologic features of the ovarian tumors included size of the largest mass and growth pattern. The size of the ovarian tumor was obtained from the gross report; however, when ovarian involvement was only microscopic, the size of the largest ovary was recorded. Foci were arbitrarily classified as nodular if carcinoma microscopically exhibited a round and well-circumscribed configuration or plaque-like if the surface of portions of non-neoplastic ovary contained tumor having a width 2 times greater than the thickness. When nodules were present, the proportion of tumor involvement in each ovary was recorded as either predominantly nodular (if ≥50% of the tumor was nodular) or a minor nodular component (if <50% of the tumor was nodular).
In cases of primary high-grade serous carcinoma of the ovary or peritoneum that did not contain a TIC on initial H&E sections, all available ovarian and tubal paraffin blocks from all non-neoplastic ovarian and fallopian tubal tissue were leveled yielding 10 additional H&E slides from each block in an effort to identify occult TIC or similar lesions involving the ovary that were not present in the original H&E slides.
Based on conventional criteria described above, carcinomas were classified as ovarian in origin in 37 cases (71%), as peritoneal in 8 (15%), and as tubal in 7 (13%) (Table 1). The only histologic type that was associated with TIC was high-grade serous carcinoma. The frequency of TIC among the high-grade serous carcinomas at each site ranged from 50% to 67% (Table 1). Of all 37 ovarian, 8 peritoneal, and 7 fallopian tube primary cases in this study, regardless of histologic type, both fallopian tubes were identified in the gross specimen in all but 4 cases. These 4 cases had fallopian tube identified only on 1 side. Three were ovarian primaries by conventional criteria and had TIC in the fallopian tube that was present. The fourth case was a peritoneal primary by conventional criteria, and TIC was not present in the fallopian that was present.
Among the 41 high-grade serous carcinomas classified as ovarian or peritoneal in origin per conventional criteria (Fig. 4), TIC was identified on the original H&E slides in 22 cases (54%). Extensive sectioning of the paraffin blocks was performed for the 19 cases in which TIC was not identified on the original H&E slides (no blocks available for 2 cases). Of these, 2 cases (both ovarian primaries based on conventional criteria) contained TIC on deeper sections, yielding a total of 24 cases with TIC (24 of 41; 59%). Intraepithelial carcinoma was not identified in the ovaries in any of these cases. Twenty-one TIC cases were unilateral, and 3 were bilateral. Of the 21 unilateral cases, 19 had TICs in the fimbria only, and 2 had TICs in the ampulla only. All TICs were in the fimbria in the 3 bilateral cases. Seventeen cases with TIC also had invasive carcinoma involving the mucosa in the same fallopian tube whereas 7 cases with TIC lacked an associated invasive carcinoma in the same tube. In the unilateral TIC cases, 15 women had bilateral involvement of the ovaries. Five women with unilateral TIC had unilateral and ipsilateral involvement of the ovary (the contralateral ovary was absent in 3 of these cases). The 21st case with a unilateral TIC (also with invasive carcinoma involving the same fallopian tube) was from a patient with primary peritoneal high-grade serous carcinoma based on conventional criteria, but the ovaries were uninvolved. In the 3 bilateral TIC cases, both ovaries were involved in each case. Thus, of all 24 cases with TIC, ovarian involvement (whether unilateral or bilateral) was ipsilateral to TIC in 23 (96%). Of TIC(−) cases with ovarian involvement (n=17), invasive carcinoma was present in the fallopian tube mucosa on the side ipsilateral to ovarian involvement (whether unilateral or bilateral) in 6 (35%).
Thirty-three high-grade serous carcinomas were classified as ovarian primaries by conventional criteria, of which 20 (61%) had TIC (Fig. 1). TIC was unifocal in 9 cases and multifocal in 11 (in each of the 3 bilateral TIC cases, intraepithelial carcinoma was multifocal in both tubes). The ovarian tumor was unilateral in 5 (15%) and bilateral in 28 (85%). The mean size of the largest ovarian tumor per patient was 6.3 cm (range: 1.6 to 17.0 cm). The growth pattern of the ovarian tumor was predominantly nodular in 9 cases (27%), nodular as a minor component in 13 cases (39%), and in the form of surface plaques in 13 cases (39%).
Eight high-grade serous carcinomas were considered peritoneal primaries based on conventional criteria, of which 4 (50%) had TIC (Fig. 2). TIC was unifocal in 1 case and multifocal in 3 cases. Ovarian involvement occurred in 7 cases, of which 2 (29%) were unilateral and 5 (71%) were bilateral. The ovaries were uninvolved in the eighth case, which also had TIC. The mean size of the largest ovary in cases with ovarian involvement was 3.6 cm (range: 2.0 to 6.0 cm). None of the cases with ovarian involvement had a predominantly nodular pattern in the ovaries, 1 (14%) had a nodular pattern as a minor component, and 5 (71%) had a surface plaque pattern of ovarian involvement.
Six high-grade serous carcinomas were classified as tubal origin per conventional criteria (Fig. 5), of which 4 (67%) had TIC (Fig. 3) (3 unilateral and 1 bilateral). Three cases had TICs in the fimbria only whereas 1 had TIC in the ampulla only. TIC was unifocal in 1 case and multifocal in 3 cases. Invasive carcinoma involved the tubal mucosa in the same fallopian tube as the TIC in all 4 cases. Of all 6 carcinomas, 4 (67%) had ovarian involvement (3 unilateral and 1 bilateral) and 2 (33%) had uninvolved ovaries. One woman with unilateral TIC had unilateral and ipsilateral involvement of the ovary. In the other 2 unilateral TIC cases, the ovaries were uninvolved. The 1 patient with bilateral TICs had involvement of both ovaries. Therefore, of the 4 TIC cases, only 2 (50%) had ovarian involvement (whether unilateral or bilateral) ipsilateral to TIC. Both TIC(−) cases with ovarian involvement had invasive carcinoma present in the fallopian tube mucosa on the side ipsilateral to ovarian involvement. The mean size of the largest ovarian tumor/ovary in cases with ovarian involvement was 3.5 cm (range: 3.0 to 4.0 cm). The growth pattern of these ovarian tumors was predominantly nodular in 2 cases (50%), nodular as a minor component in 1 (25%), and in the form of surface plaques in 1 (25%).
A comparison of pathologic features of high-grade serous carcinomas with and without TIC is shown in Table 2. Statistical analysis was not performed because of the relatively small number of cases in each subgroup; however, primary ovarian/peritoneal high-grade serous carcinomas with or without TICs had similar size of the largest ovarian tumor, frequency of bilaterality of ovarian involvement, and frequency of various growth patterns (Figs. 6, 7) within the ovary. Similar observations of primary fallopian tube high-grade serous carcinomas were noted for these same parameters when comparing cases with and without TIC.
Based on conventional criteria, 70%, 17%, and 13% of high-grade serous carcinomas qualified as primary ovarian, peritoneal, and tubal origin, respectively. Using the presence of TIC to designate tubal origin, in addition to conventional criteria for cases lacking TIC, 28%, 8%, and 64% of cases were classified as ovarian, peritoneal, and tubal primary origin, respectively (Fig. 8). This reassignment, based on the inclusion of TIC as a supplemental criterion, resulted in a 51% increase in tumors classified as tubal primaries.
The results of this study provide strong support that TIC is specifically associated with serous rather than nonserous carcinomas as proposed by Roh et al,19 who found that TIC was associated with serous but not endometrioid pelvic carcinomas. Accordingly, a more appropriate designation is serous TIC (STIC) as previously suggested.12
Our study also confirms that complete examination of all fallopian tube tissue in cases of pelvic (nonuterine) serous carcinoma identifies STIC in 61% of ovarian high-grade serous carcinomas, which is higher than the 48% reported by Kindelberger et al.14 The higher frequency may, in part, be due to our leveling of all the paraffin blocks from the tube in which a STIC was not initially identified, which resulted in the detection of 2 additional STICs. The finding of STIC with simultaneous high-grade serous carcinoma involving the ovary or peritoneum can be interpreted in 3 ways: (1) primary fallopian tube carcinoma with secondary involvement of the ovary/peritoneum, (2) the primary ovary or peritoneal serous carcinoma with secondary involvement of the fallopian tube in the form of intraepithelial spread, and (3) independent primary serous carcinomas of the involved sites.
In this study, TIC was frequently associated with invasive carcinoma in the mucosa of the ipsilateral fallopian tube, a substantial number of TIC(−) cases had invasive carcinoma in the tubal mucosa on the side ipsilateral to the ovarian tumor, and ovaries in cases with TIC were slightly more often nodular compared with those without TIC. Based on the findings in this study, and of others,7,12,14,17,19,20 the weight of evidence supports the interpretation that the site of origin is the fallopian tube in view of the following: (1) the presence of STICs in isolation in cases with or without a genetic predisposition to ovarian carcinoma,1,4,5,8,16,22 (2) the fimbriated end of the tube is the preferred site of STIC and that the fimbriated end is in close proximity to the ovarian surface/peritoneal cavity,18 (3) high frequency of STICs in tubal primary carcinomas (using conventional criteria), especially in the fimbriated end of the tube, in our study and in another study,14 (4) the presence of STICs or p53 signatures (the latter defined as normal appearing nonciliated tubal epithelium that shows diffuse and strong immunohistochemical expression of p53) with TP53 mutations in the absence of an associated ovarian carcinoma,17 (5) the presence of a STIC and p53 signature in the same case with identical TP53 mutations,17 (6) the finding of fallopian tube epithelial atypia/dysplasia in isolation exhibiting aneusomy for multiple chromosomes,20 and (7) the presence of identical TP53 mutations in STICs/p53 signatures and synchronous ovarian/peritoneal high-grade serous carcinomas.7,17
Although mounting evidence suggests that ovarian high-grade serous carcinoma may actually originate from the fallopian tube and that STICs are the likely precursor, it is important to note that not all of our cases were associated with a STIC. On initial sections, 45% of our high-grade serous carcinomas classified as primary ovarian origin using conventional criteria did not have a STIC. Even after further sectioning of the paraffin blocks, which revealed 2 additional cases with STIC (10% of all ovarian primaries associated with STIC), 39% of ovarian primaries were still unassociated with STICs. Also, 24% of cases had neither intraepithelial nor invasive carcinoma in tubal mucosa (data not shown). In the study by Kindelberger et al,14 a STIC was not found in 22 of 42 (52%) cases classified as primary ovarian serous carcinoma, and 29% of their cases also lacked intraepithelial and invasive carcinoma in tubal mucosa. Although this suggests that a subset of high-grade serous carcinomas might arise directly from the ovarian surface epithelium as has been previously thought, another plausible explanation is that normal tubal epithelium can implant on the ovary at the time of ovulation when the surface ovarian epithelium is ruptured resulting in the formation of an inclusion cyst which can subsequently undergo malignant transformation.15 This idea is supported by the finding that epithelial cells from tubal mucosa are easily shed after flushing the fallopian tube.15 Other possibilities which may explain the origin of ovarian high-grade serous carcinoma include: (1) the rare serous borderline tumors or low-grade serous carcinomas that have undergone high-grade transformation and (2) malignant transformation of ovarian cortical inclusion glands via a metaplastic process from the surface epithelium as opposed to implantation of tubal mucosa.15
Of note, TIC and tubal “carcinoma in situ” have been described in cases of tubal malignant mesodermal mixed tumor (also, in our experience; unpublished data).6,9 Hence, STIC may also play a role in the development of “ovarian” malignant mesodermal mixed tumor with a serous carcinoma component in addition to pure high-grade serous carcinoma.
The conventional criteria that have been used to assign primary site of origin for ovarian, peritoneal, and tubal carcinomas3,11,21 are arbitrary and have never been validated. It now appears that the presence of a STIC is a useful criterion for establishing primary tubal versus ovarian and peritoneal origin. In this regard, we also evaluated whether bilaterality, tumor size, nodular growth pattern, and surface plaques of tumor involving the ovary were associated with STIC and, thus, if they were useful in assigning primary tubal origin, but these features were not helpful (Table 2). As the presence of a STIC has not been previously used as a uniformly accepted criterion for the diagnosis of tubal carcinoma, prior data from tumor registries and clinicopathologic studies have underestimated the actual frequency of primary fallopian tube carcinoma and, conversely, overestimated the frequency of primary ovarian and peritoneal carcinomas. On the basis of population-based data, the frequency distribution of primary ovarian, peritoneal, and tubal serous carcinomas is 78% to 90%, 8% to 16%, and 2% to 6%, respectively.10,13 Similarly, in our study, the distribution of primary ovarian, peritoneal, and tubal high-grade serous carcinomas using conventional criteria was 70%, 17%, and 13%, respectively; however, if STIC is used as a supplemental criterion to define tubal origin, then these figures would have changed to 28%, 8%, and 64%, respectively. A substantial shift in the frequency distribution of primary site of origin based on the presence of STIC was also observed in another study.12
In conclusion, TIC is frequently associated with tumors classified as ovarian/peritoneal high-grade serous carcinoma using conventional criteria, and this association is specific for high-grade serous carcinoma as opposed to other histologic types, supporting the designation of the lesion as STIC. In contrast, other features of ovarian tumors that suggest metastatic disease, such as bilaterality, small size, nodular growth pattern, and surface plaques, are not predictive of tubal origin. Although most ovarian/peritoneal high-grade serous carcinomas might originate from the fallopian tube, a proportion may arise from inclusion cysts/surface epithelium which we postulate may have developed from implantation of normal tubal epithelium on the ovary.15 Thus, it is conceivable that all high-grade pelvic serous carcinomas could arise from tubal epithelium. Additional molecular genetic studies are necessary to establish that STIC is the earliest form of carcinoma rather than intraepithelial spread from adjacent invasive serous carcinoma of ovarian or peritoneal origin. Determination of the primary site (ovarian, peritoneal, or tubal) of pelvic high-grade serous carcinomas does not currently alter standard therapy; however, the data implicating the fallopian tube as the primary site of high-grade pelvic serous carcinoma will, in the future, have an important implication for the development of new approaches for early detection, treatment, and prevention of this highly lethal disease.
1. Agoff SN, Garcia RL, Goff B, et al. Follow-up of in situ and early-stage fallopian tube carcinoma in patients undergoing prophylactic surgery for proven or suspected BRCA-1 or BRCA-2 mutations. Am J Surg Pathol. 2004;28:1112–1114.
2. Bell DA. Origins and molecular pathology of ovarian cancer. Mod Pathol. 2005;18(suppl 2):S19–S32.
3. Bloss JD, Liao SY, Buller RE, et al. Extraovarian peritoneal serous papillary carcinoma: a case-control retrospective comparison to papillary adenocarcinoma of the ovary. Gynecol Oncol. 1993;50:347–351.
4. Callahan MJ, Crum CP, Medeiros F, et al. Primary fallopian tube malignancies in BRCA-positive women undergoing surgery for ovarian cancer risk reduction. J Clin Oncol. 2007;25:3985–3990.
5. Carcangiu ML, Peissel B, Pasini B, et al. Incidental carcinomas in prophylactic specimens in BRCA1 and BRCA2 germ-line mutation carriers, with emphasis on fallopian tube lesions: report of 6 cases and review of the literature. Am J Surg Pathol. 2006;30:1222–1230.
6. Carlson JA Jr, Ackerman BL, Wheeler JE. Malignant mixed mullerian tumor of the fallopian tube. Cancer. 1993;71:187–192.
7. Carlson JW, Miron A, Jarboe EA, et al. Serous tubal intraepithelial carcinoma: its potential role in primary peritoneal serous carcinoma and serous cancer prevention. J Clin Oncol. 2008;26:4160–4165.
8. Colgan TJ, Murphy J, Cole DE, et al. Occult carcinoma in prophylactic oophorectomy specimens: prevalence and association with BRCA germline mutation status. Am J Surg Pathol. 2001;25:1283–1289.
9. Gagner JP, Mittal K. Malignant mixed Mullerian tumor of the fimbriated end of the fallopian tube: origin as an intraepithelial carcinoma. Gynecol Oncol. 2005;97:219–222.
10. Goodman MT, Shvetsov YB. Incidence of ovarian, peritoneal, and fallopian tube carcinomas in the United States, 1995-2004. Cancer Epidemiol Biomarkers Prev. 2009;18:132–139.
11. Hu CY, Taymor ML, Hertig AT. Primary carcinoma of the fallopian tube. Am J Obstet Gynecol. 1950;59:58–67.
12. Jarboe E, Folkins A, Nucci MR, et al. Serous carcinogenesis in the fallopian tube: a descriptive classification. Int J Gynecol Pathol. 2008;27:1–9.
13. Jordan SJ, Green AC, Whiteman DC, et al. Serous ovarian, fallopian tube and primary peritoneal cancers: a comparative epidemiological analysis. Int J Cancer. 2008;122:1598–1603.
14. Kindelberger DW, Lee Y, Miron A, et al. Intraepithelial carcinoma of the fimbria and pelvic serous carcinoma: evidence for a causal relationship. Am J Surg Pathol. 2007;31:161–169.
15. Kurman RJ, Shih I. The origin and pathogenesis of epithelial ovarian cancer: a proposed unifying theory. Am J Surg Pathol. 2010;34:433–443.
16. Lamb JD, Garcia RL, Goff BA, et al. Predictors of occult neoplasia in women undergoing risk-reducing salpingo-oophorectomy. Am J Obstet Gynecol. 2006;194:1702–1709.
17. Lee Y, Miron A, Drapkin R, et al. A candidate precursor to serous carcinoma that originates in the distal fallopian tube. J Pathol. 2007;211:26–35.
18. Medeiros F, Muto MG, Lee Y, et al. The tubal fimbria is a preferred site for early adenocarcinoma in women with familial ovarian cancer syndrome. Am J Surg Pathol. 2006;30:230–236.
19. Roh MH, Kindelberger D, Crum CP. Serous tubal intraepithelial carcinoma and the dominant ovarian mass: clues to serous tumor origin? Am J Surg Pathol. 2009;33:376–383.
20. Salvador S, Rempel A, Soslow RA, et al. Chromosomal instability in fallopian tube precursor lesions of serous carcinoma and frequent monoclonality of synchronous ovarian and fallopian tube mucosal serous carcinoma. Gynecol Oncol. 2008;110:408–417.
21. Sedlis A. Carcinoma of the fallopian tube. Surg Clin North Am. 1978;58:121–129.
22. Shaw PA, Rouzbahman M, Pizer ES, et al. Candidate serous cancer precursors in fallopian tube epithelium of BRCA1/2 mutation carriers. Mod Pathol. 2009;22:1133–1138.
23. Vang R, Shih I, Kurman RJ. Ovarian low-grade and high-grade serous carcinoma: pathogenesis, clinicopathologic and molecular biologic features, and diagnostic problems. Adv Anat Pathol. 2009;16:267–282.