BACKGROUND:Serous tubal intraepithelial carcinoma (STIC) and the p53 signature in tubal mucosa have been supported to be precursor lesions in high-grade serous carcinoma (HGSC) of the fallopian tube, ovary, and peritoneum. It remains critical to find biomarkers for precursor lesions in order to detect HGSCs efficiently. IMP3 is an oncoprotein that has been explored in human malignancies. No studies have specifically addressed the expression of IMP3 in precursor or early lesions of HGSC. The main purposes of this study are to evaluate if IMP3 plays any role in the process of pelvic serous carcinogenesis by examining its expression in HGSC precursor lesions, to examine the relationship between IMP3 and p53 in those precursor lesions, and to check if IMP3 can be used as a biomarker for early diagnosis.METHODS:Immunohistochemistry for IMP3 and p53 was performed and evaluated in 48 HGSCs with STIC, 62 HGSCs without STIC, and 60 benign cases as negative controls. Sections of fallopian tubes with or without STIC , as well as cancers within the ovaries, were studied. IMP3 signature was defined as strong IMP3 cytoplasmic staining in 10 or more consecutive benign-looking tubal epithelial cells. The relationship between IMP3 and p53 overexpression was examined.RESULTS:In the 48 HGSC patients with STIC, IMP3 was positive in 46% of STIC lesions and had a similar positive rate in the invasive components of HGSC. IMP3 was also expressed in normal appearing tubal epithelia (IMP3 signature) in 15 (31%) of 48 HGSC cases with STIC and 10 (16%) of 62 cases without STIC. In contrast, no single IMP3 signature was found in the benign control group. Concordant expression of IMP3 and p53 signatures in the STIC group was found in up to one-third of the cases. There were also five (10%) STIC cases with positive IMP3 and negative p53.CONCLUSIONS:We conclude that IMP3 may be involved in the process and progression of pelvic HGSC and may serve as a complimentary biomarker in diagnosing STIC.

Background: Mounting evidence suggests the fallopian tube as the origin for ovarian high grade serous carcinoma (HGSC). We attempted to identify the tubal cytological features that allow us to distinguish malignant from benign conditions. Methods: Tubal specimens (n = 56) were collected from patients who underwent bilateral salpingo-oophorectomy (BSO) due to various clinical indications. A standard procedure to collect fallopian tube brushings from freshly received surgical specimens was developed. Cytological diagnoses were classified into three categories: benign, atypical, and suspicious for malignancy/malignant. Cytological variables of individual cells and epithelia were subjected to statistical analysis. The fallopian tube histology was used as diagnostic reference for confirmation of cytology diagnosis. Results: Among the 56 fallopian tube specimens, 2 (3.7 %) showed inadequate cellularity preventing further evaluation, 11 (20.4 %) were diagnosed as malignant or suspicious of malignancy, 7 were atypical, and 36 were benign. The presence of three dimensional clusters (p < 0.0001, Fisher's Exact Test), or prominent nucleoli (p = 0.0252, Fisher Exact test) was highly correlated with the diagnosis of malignancy. The suspicious malignant/malignant cytological diagnosis was also highly correlated with presence of HGSC with or without serous tubal intraepithelial carcinoma (STIC). Conclusions: Tubal cytology may be useful for ovarian cancer screening and early detection.