Tumor response to NAC predicts survival and can be considered a surrogate prognostic marker. Three tiered chemotherapy response score (CRS) of omental tissue sections showed a significant association with survival. In patients with CRS 1 or 2, NAC selects a subpopulation of chemotherapy resistant tumor cells. This study will examine comprehensive molecular analyses on the residual disease of 104 clinically defined high-grade serous carcinoma after NAC, including next-generation sequencing on 14 matched pretreatment biopsies. This information together with immune marker expression and BRCA expression, will provide a unique opportunity to guide biomarker-driven adjuvant studies targeting these chemotherapy-resistant tumor cells.

DNA extracted from FFPE block will be used. Capture sequencing and immunohistochemistry will be performed.

Immunohistochemistry The formalin-fixed, paraffin-embedded tissue blocks were sectioned at 4 µm thickness onto Superfrost Plus glass slides (Thermo Fisher Scientific, Waltham, MA, USA). These sections were deparaffinized in xylene and rehydrated through graded alcohols. Immunohistochemical staining was performed using an automatic immunostaining instrument (Ventana Benchmark XT; Ventana Medical Systems), according to the manufacturer's recommendations. Antigen retrieval was performed using a Cell Conditioning Solution (CC1; Ventana Medical Systems). The sections were incubated with antibodies against PD-L1 (prediluted, clone SP263, Ventana Medical Systems), CD8 (prediluted, clone C8/144B, Dako, Glostrup, Denmark), PD-1 (1:50, clone NAT105, Cell Marque, Rocklin, CA, USA), Foxp3 (1:50, clone 236A/E7, Abcam, Cambridge, UK), ICOS/CD278 (1:50, clone SP98, Thermo Fisher Scientific), and LAG-3 (1:100, clone EPR4392(2), Abcam). After chromogenic visualization using an ultraView Universal DAB Detection Kit (Ventana Medical Systems), the slices were counterstained with hematoxylin, dehydrated in graded alcohols and xylene, and then embedded in mounting solution. Appropriate positive and negative controls were concurrently stained to validate the staining method.

Targeted Capture Sequencing on NextSeq550 using SureSelect XT Sample Preparation DNA was extracted from microdissected tumor rich areas from formalin fixed paraffin embedded tumor sections using DNAeasy kit (QIAGEN, Venlo, Netherlands). Two hundred fifty nanogram of genomic DNA in 50μl TE buffer was fragmented to a median size of 150bp using the Covaris-E220 AFA instrument (Covaris, Woburn, MA) with the following settings: peak incident power 175 watts, duty factor 10.0%, cycles per burst 200 cycles, run time 500 sec at 4oC. DNA fragments were evaluated using capillary electrophoresis on High Sensitivity D1000 ScreenTape and Reagents (TapeSation4200, Cat.No.5067-5584, 5585; Agilent Technologies, Santa Clara, CA, USA). End repair, 3'-end adenylation and sequencing adapter ligation of DNA fragments were performed following the manufacture's protocol (SureSelect XT Reagent kit, HSQ Cat.No.G9611A, G9611B; Agilent Technologies, Santa Clara, CA, USA). After each steps, fragments were purified using AMPure XP beads (Cat. No. A63382; Beckman Coulter, High Wycombe, UK). The resulting DNA was amplified by 14 cycles PCR amplification (SureSelect Herculase Ⅱ Fusion Enzyme with dNTP Combo 200 RXN kit, Cat. No. 600677; Agilent Technologies, Santa Clara, CA, USA). The quality of the PCR product was assessed by capillary electrophoresis with DNA1000 TapeScreen and Reagents (TapeSation4200, Cat.No.5067-5582, 5583; Agilent Technologies, Santa Clara, CA, USA) and the concentration was measured by Qubit 2.0 Fluorometer (Qubit dsDNA HS Assay Kit, Cat. No. Q32854; Thermofisher Scientific, Inc, Waltham, MA).

Capture and Enrichment Target enrichment was performed according to the manufacturer's instructions (SureSelect XT Custom 0.5Mb-2.9Mb library, Cat. No. 5190-4816, 4817; Agilent Technologies, Santa Clara, CA, USA). The resulting captured libraries with indexing primers were amplified by 12 cycles of PCR (SureSelect Herculase Ⅱ Fusion Enzyme with dNTP Combo 200 RXN kit, Cat. No. 600677; Agilent Technologies, Santa Clara, CA, USA) and purified again. Quantity and quality of indexed library DNA were verified by capillary electrophoresis with High Sensitivity D1000 ScreenTape and Reagents (TapeSation4200, Cat.No.5067-5584, 5585; Agilent Technologies, Santa Clara, CA, USA) and Qubit 2.0 Fluorometer (Qubit dsDNA HS Assay Kit, Cat. No. Q32854; Thermofisher Scientific, Inc, Waltham, MA).

Sequencing We diluted resulting libraries to 4nmol/l and pooled by combining 5μl of each diluted library for normalization. Subsequently, pooled library was denatured into single strands by using fresh 0.2N NaOH and 200mM Tris-HCl, pH7. A PhiX control was denatured in the same way. The concentration of final loading library was 1.8pmol/l and Phix was spiked in at 1%. A standard flow cell was loaded on the Illumina NextSeq550 and sequencing was conducted with 2X100bp paired-end reads using NextSeq550 reagent v2 kit with high-output 300 cycles. (NextSeq550 System user guide part #15069765-V02 Mar 2016, NextSeq550 Reagent V2 Kit High-output 300 cycles Cat. No. FC-404-2004; Illumina, CA).

Inclusion Criteria:

• patients with pathologically confirmed epithelial ovarian cancer who received at least 1 cycle of neoadjuvant chemotherapy
• The patient inclusion criteria included 1) high grade serous carcinoma 2) stage III/IV disease 3) availability of clinical data and tumor tissue 4) viable residual tumors after neoadjuvant chemotherapy.

Exclusion Criteria:

• patients with complete pathologic response (CRS 3) as IHC staining was insufficient or unavailable

tirtiary medical center