Comparison of predictive models in women with advanced epithelial ovarian cancer triaged to primary debulking or neoadjuvant chemotherapy.

Abstract

5547 Background: Triaging women with advanced epithelial ovarian cancer (AEOC) into primary debulking surgery (PDS) or neoadjuvant chemotherapy (NACT) remains largely subjective. Methods: This prospective observational study recruited women over 18 years with stage III-IV AEOC. Decision for PDS or NACT was based on patient-specific factors and radiological parameters. The agreement between clinical decisions and predictive models—Aletti’s surgical complexity score, MSKCC Team Ovary criteria, Mayo triage algorithm, and Integrated Predictive Model (IPM) was assessed using kappa statistics. Results: 72 women with AEOC were included, with 17 (23.6%) assigned to the PDS and 55 (76.4%) to NACT. Amongst NACT patients, interval debulking surgery (IDS) was feasible in 30 women (54.5%), while 25 (45.5%) did not undergo surgery due to reasons such as disease progression, death, poor performance status, stable disease, or loss to follow-up. No difference was observed between the NACT and PDS groups in demographic parameters. Performance scores differed significantly, with higher scores observed in the NACT group compared to the PDS group: median ECOG (2 [1–2] vs. 1 [1–1], p < 0.001), ASA score (2 [2–3] vs. 2 [2–2], p = 0.005), and frailty index (0.26 ± 0.11 vs. 0.15 ± 0.05, p < 0.001). Serum albumin levels were lower (3.0 ± 0.51 vs. 3.7 ± 0.30 g/dL, p < 0.001), and median CA125 levels were higher (1170 [341–2637] vs. 494 [219.7–1000] U/mL, p < 0.001) in the NACT group. Radiological parameters, including median peritoneal carcinomatosis index (PCI) scores (16 [10–23] vs. 5 [3–8], p < 0.001), volume of ascites, pleural effusion, and disease at challenging operative sites, were also higher in the NACT group (p < 0.05). Surgical outcomes, including surgical PCI scores (5[2-6] vs 6[3-11], =0.35), residual disease rates (complete/optimal debulking: 96.6% vs. 88.2%, p = 0.42), surgical complexity score (4.7 ± 1.45 vs. 4.4 ± 1.33, p = 0.82), blood transfusion rates (80% vs. 70.6%, p = 0.76), and grade 2-3 complications (60% vs. 58.8%, p = 0.58) were similar in both groups. Baseline predictive scores were significantly higher in the NACT group compared to the PDS group: Aletti’s surgical complexity score (8.4 ± 2.80 vs. 5.2 ± 1.25, p < 0.001), MSKCC Team Ovary criteria (6.4 ± 3.31 vs. 1.9 ± 1.49, p < 0.001), Mayo triage algorithm (0.87 ± 0.39 vs. 0.24 ± 0.44, p < 0.001), and IPM final score (high-risk: 69.1% vs. 52.8%, p < 0.001). Clinical decisions showed moderate concordance with the Mayo triage algorithm (κ = 0.57) and IPM score (κ = 0.51) and fair concordance with Aletti’s score (κ = 0.33) and MSKCC criteria (κ = 0.23). Conclusions: Triage decisions based on patient performance status, nutritional factors, and disease extent demonstrated moderate concordance with predictive models. Both PDS and IDS had excellent cytoreductive outcomes with similar perioperative performance aligning with results from literature.