Ovarian Tumors for Tissue Microarray (TMA) Construction. A TMA containing 540 ovarian tumor samples from 540 patients was con­structed. The average age of the patients in the co­hort was 55.25 years. There were 278 malignant tu­mors, 39 low malignant potential tumors (LMP) and 223 benign tumors (Table 1). The histological vari­ants were: 248 serous, 61 mucinous, 85 endometri-oid, two clear cell tumors, two Brenner tumors, 20 mixed epithelial tumors, 31 undifferentiated, 21 un­classified and 50 non epithelial tumors (granulosa-cell, germline-cell). The slides from all tumors were previously reviewed by a single pathologist. Tumor stage and grade were defined according to the Inter­national Federation of Gynecology and Obstetrics

(FIGO) and World Health Organization (WHO) cri­teria [11].

Investigated Bacterial Artificial Chromo­some Clones from 1q and Genes They Contain. We investigated the BAC of 1q21.3-RP11-126K1, of 168,686 bp (http://genome.ucsc.edu/). This contains the following genes: PIP5K1A; RFX5; PIK4CB; SELENBP1, which encode for phospha-ditylinositol-4-phosphate-5-kinase a, RFX5 protein (involved in expression of MHC-II), phosphadityl-inositol-3-kinase, and Selen-binding protein 1, re­spectively.

We also investigated the BAC of 1q23.3-RP11-97G24, of 171,855 bp. It contains the genes SLAMF1; SLAMF6; CD84, which encode for sig­naling lymphocytic activation molecule family member 1 (involved in cytolysis activity of NK cells), signaling lymphocytic activation molecule family member 6, and protein of super family of immunoglobulin receptors, respectively.

Tissue Microarray. For TMA construction, a hematoxilin and eosin (H&E)-stained section was made from each block to define representative tumor areas. Tissue cylinders with a diameter of 0.6 mm were punched from tumor areas and brought into a recipient paraffin block using a custom-made pre­cision instrument [12]. Samples were distributed in

one regular-sized recipient paraffin block contain­ing 540 specimens. Five pm sections of the blocks were transferred to glass slides using a paraffin-sec­tioning aid system (adhesive coated slides, adhesive tape, UV-lamp; Instrumedics Inc., Hackensack, NJ, USA).

Fluorescence in situ Hybridization (FISH).

Prior to hybridization the slides were treated with xylene at 37°C, absolute ethanol at room tempara-ture and 1 M sodium thiocyanate at 80° C for depar-rafinization and pretreatment. Dual color FISH was performed using the locus RP11-126K1 (1q21.3) clone probe, conjugated with digoxigenin and de­tected in red, and locus RP11-97G24 (1q23.3) clone probe, conjugated with biotin and detected in green. Labeling of the probes was performed by nick-trans­lation. Denaturation of the DNA was carried out at 75°C for 10 min. (probe mixture) or at 73°C for 5 min. (slides). The probe mixture was applied to the slides and hybridized overnight in a moist cham­ber at 37°C. The post hybridization washes were performed using formamide 50%/2xSSC. A three-step detection was achieved by blocking non spe­cific signals with block buffer, first detection with FITC-avidin/ mouse anti-Dig, again blocking, sec­ond detection with anti-avidin/Dig anti-mouse IgG, blocking, and third detection with FITC-avidin/ anti-Dig rhodamine. The slides were counterstained with DAPI in antifade. The presence of more than five locus signals per cell or tight clusters in at least 10% of tumor cells was considered to connote am­plification. The presence of 3-5 locus signals in at least 10% of tumor cells was considered to connote a gain.

Statistical Analysis. The relationship between copy number changes and clinicopathological data was estimated using the %2 test and the p value was calculated. A p value of <0.05 was required for sig­nificance.