Model Including Diffusion-Weighted MRI More Accurately Predicts Ovarian Cancer Resectability

A modified predictive model combining the Memorial Sloan Kettering Cancer Center (MSKCC) model with MRI features showed improved accuracy in predicting ovarian cancer resectability compared with the MSKCC model alone. The findings, which were published in the American Journal of Obstetrics & Gynecology, suggest the modified model is useful for predicting primary debulking surgery outcomes and lay the groundwork for larger cohort studies to develop tailored treatment strategies.

In patients with high-grade serous ovarian cancer (HGSOC), complete resection during primary debulking surgery is correlated with favorable prognosis, and assessing resectability accurately ahead of surgery is an important part of individualized HGSOC management. With accurate assessment, clinicians can identify patients who may experience the most benefit from either primary debulking surgery or neoadjuvant chemotherapy, the study authors explained.

The widely used, noninvasive model created at MSKCC assesses resectability based on 3 preoperative clinical criteria and 8 radiologic criteria. The modified model assessed in the new study utilizes the 8 radiologic criteria in the MSKCC model plus diffusion-weighted MRI (DW-MRI)–depicted growth patterns of metastatic lesions in the subcapsular or diaphragm and mesenteric regions. DW-MRI is particularly useful and highly sensitive in identifying abdominal metastases and implants, especially in the diaphragm and small intestine.

A total of 184 patients with HGSOC who received DW-MRI prior to primary debulking surgery at 2 medical centers were included in the study. The patients were separated into 3 cohorts: a study cohort (n = 100), an internal validation cohort (n = 46), and an external validation cohort (n = 38), and preoperative radiologic evaluations were conducted by 2 radiologists independently using both the MSKCC model and the DW-MRI­–based modified model.

MRI machine | Image credit: phonlamaiphoto - stock.adobe.com

Scores based on both models showed excellent intraobserver and interobserver agreement, with the MSKCC model exhibiting intraclass correlation coefficients of 0.980 and 0.959, respectively (P < .001), and the DW-MRI–based models showing intraclass correlation coefficients of 0.962 and 0.940, respectively (P < .001). Both models also independently predicted surgical resectability, with ORs of 1.825 (95% CI, 1.390-2.395; P < .001) and 1.776 (95% CI, 1.410-2.238; P < .001) using the MSKCC model and the modified model.

Predictive performance was improved with the DW-MRI–based model, which had an area under the curve (AUC) of 0.867 in the study cohort, 0.806 in the internal validation cohort, and 0.913 in the external validation cohort. The MSKCC model exhibited an AUC of 0.700 in the internal validation cohort (P = .004) and 0.819 in the external validation cohort (P = .013). Patients with scores of 0 to 2, 3 to 4, 5 to 6, 7 to 10, and 11 or higher within the modified model experienced complete debulking rates of 90.3%, 66.7%, 53.3%, 11.8%, and 0%, respectively. Those with incomplete tumor debulking mostly had infiltrative tumors.

“Given that many residual lesions tend to be found in the diaphragm and mesentery, which are not included in the MSKCC model, our study modified the model by incorporating 2 clinically significant sites and the growth patterns of lesions as criteria, thereby demonstrating enhanced predictive performance of the resectability of HGSOC,” the authors wrote.

The study was limited by a small sample size despite including internal and external validation cohorts, as well as by its retrospective nature. Still, the authors noted that it provided valuable insight and provided a framework for future prospective, multicenter studies with larger cohorts to explore the value of DW-MRI in predicting resectability in HGSOC.

Reference

Lu J, Guo Q, Zhang Y, et al. A modified diffusion-weighted magnetic resonance imaging–based model from the radiologist’s perspective: improved performance in determining the surgical resectability of advanced high-grade serous ovarian cancer. Am J Obstet Gynecol. Published online March 1, 2024. doi:10.1016/j.ajog.2024.02.302