Low-grade Serous Carcinoma of the Ovary: Clinicopathologic Analysis of 52 Invasive Cases and Identification of a Possible Noninvasive Intermediate Lesion. American journal of surgical pathology Ahn, G., Folkins, A. K., McKenney, J. K., Longacre, T. A. 2016; 40 (9): 1165-1176

Abstract

Low-grade serous carcinoma (LGSC) is an uncommon but distinct histologic subtype of ovarian carcinoma. Although the histologic features and natural history of LGSC have been described in the literature, there is no robust correlative study that has specifically addressed histologic features in correlation with clinical follow-up. To refine the criteria for invasion patterns of LGSC and determine additional clinically pertinent morphologic features of LGSC predisposing to a more aggressive clinical course, the clinicopathologic features of 52 LGSCs were evaluated and compared with those of a large series of serous borderline tumors (SBT), with and without invasive implants. To qualify for LGSC, the tumor needed to demonstrate destructive invasion, nuclear atypia that was mild to moderate at most (grade 1 or 2), and a mitotic index that did not exceed 12 mitoses per 10 high-power fields. On the basis of histologic evaluation, destructive invasion was classified into 7 primary architectural patterns: (1) micropapillary and/or complex papillary; (2) compact cell nests; (3) inverted macropapillae; (4) cribriform; (5) glandular and/or cystic; (6) solid sheets with slit-like spaces; and (7) single cells. Five-year overall survival and disease-free survival for LGSC were 82% (median, 72 mo) and 47% (median, 54 mo), respectively. All the patients with fatal outcome demonstrated tumors showing invasion with predominant patterns of cribriform glands, micropapillae and/or complex papillae, or compact cell nests. Notably, 2 of 9 patients with fatal outcome had only small foci of destructive invasion (2 and 3 mm, respectively) with compact cell nests and cribriform glands as the predominant patterns. There was no statistically significant association between pattern of invasion and disease-free survival. Classic stromal microinvasion, as defined by nondestructive stromal invasion <5 mm was identified in 52% of LGSC and was statistically more frequent in LGSC than in SBT (P<0.001). In 2 LGSCs, there were areas demonstrating an intraluminal solid proliferation of tumor cells with grade 1 or 2 nuclear atypia, which we hypothesize may represent a noninvasive form of LGSC, as similar non-invasive proliferations of morphologically low-grade serous carcinomatous cells were also identified in 8 SBTs, in either solid or compact glandular/papillary formations. One patient with this isolated noninvasive pattern in SBT developed LGSC 40 months after initial operation. LGSC was typically high stage (FIGO stages II to IV, 86%) and bilateral (68%), with multiple foci of invasion (82%). Bilaterality was significantly more common in high-stage disease (P=0.009). LGSC was associated with SBT in 84% of cases, most commonly usual type (27%), followed by cribriform (18%), micropapillary (11%), or mixed cribriform and micropapillary (7%) types; focal micropapillary and/or cribriform features were present in an additional 16%. The presence of intraluminal proliferations of cells resembling LGSC occurring in SBT should prompt additional tumor sampling and assiduous evaluation of implants (if present), as this appears to represent a form of intraepithelial carcinoma, which may be associated with invasion elsewhere.

View details for DOI 10.1097/PAS.0000000000000693

View details for PubMedID 27487741