Article Text
PDF
Abstract
Introduction/Background Endometrial cancer (EC) relapse is a heterogeneous disease whose recurrence patterns vary between loco-regional, lymph-nodal, parenchymal or peritoneal. Recurrences can occur as single or multiple nodules and follow single or mixed pathway. Only a minority of patient is considered eligible for secondary cytoreductive surgery (SCS). We retrospectively analyzed clinico-histological-radiological variables of EC relapse, and hypothesized a preoperative predictive score of complete gross resection (CGR).
Methodology Multicentric retrospective analysis including patients with recurrent EC (January-2010-December-2021). Multivariate analysis was performed to evaluate factors that could predict CGR. Each significant variable was assigned a ’predictive score’. The total predictive score of all patients was calculated and the corresponding CGR rate determined. The score was then validated using an additional small internal population.
Results Two-hundred-forty-three patients/331 (73%) were evaluated to undergo surgery. Of them, 186 (56%) received SCS, while 17.2% underwent diagnostic laparoscopy. At multivariate analysis, age<65 (OR 2.530,p=0.025), single-site relapse (OR 3.140,p=0.006), lymph-node(OR 4.363,p=0.004) and parenchymal relapse(OR 5.689,p=0.021) were confirmed as positive predictors for CGR. A value of 1 has been assigned to each significant variable. The sum formed the overall predictive risk score, which ranged between 0–3. An increasing rate in CGR was recorded going from score 0 to 3 (CGR score0 vs 3:33.3%vs93.3%) (figure 1). A cut-off of 2 (0–1 versus 2–3) was identified according to the Youden-Index, obtaining a sensitivity=64.6%, specificity=75.4%, accuracy=67.5%, positive predictive value=87.8% and negative predictive value=43.8%. The same trend was confirmed in the validation population(figure 1).
Conclusion Age<65 years, single-site relapse, nodal and parenchymal pathways were positive predictors of CGR. According to our score, an additional 20% of patients with a score ≥ 2 would have been candidates for surgery with a probability of CGR above 80%. For patients with score 1, examination under anesthesia and/or diagnostic laparoscopy could be considered as useful tools to complete surgical feasibility assessment.
Disclosures None