Abstract / Summary
Objective: This retrospective study tested the value of magnetic resonance imaging (MRI) dynamic enhancement and diffusion-weighted imaging (DWI) for evaluating villous vascular endothelial growth factor (VEGF) expression in early tubal pregnancy. The biological activity of the trophoblast is relevant to therapeutic planning, but no first-line radiological study currently quantifies it before treatment. Methods: This study retrospectively analyzed 50 patients with tubal pregnancy who underwent preoperative MRI, including dynamic enhancement and DWI from August 2021 to December 2023 in our hospital. Of 81 consecutively identified patients with clinically diagnosed tubal pregnancy who consented to both MRI and surgical treatment, 31 were excluded at a second assessment performed before MRI (amenorrhea >70 days, n = 16; previous EP-directed drug treatment, n = 12; contraindications to MRI, n = 3); the remaining 50 underwent preoperative MRI followed by surgery and constitute the analyzed cohort (Figure 1). The apparent diffusion coefficient (ADC) values and the maximum enhancement amplitude for the gestational sac (GS) wall were measured. Vascular endothelial growth factor (VEGF) expression in the villi specimens was determined by immunohistochemical labeling. The association analysis was carried out between ADC values or the maximum enhancement amplitude and VEGF expression levels in villi specimens. Results: By Spearman rank correlation, the maximum enhancement amplitude correlated negatively with ADC values of the villi (ρ = −0.655, 95% CI -0.852 to −0.376, p < 0.001) and positively with VEGF expression levels (ρ = 0.791, 95% CI 0.578 to 0.934, p < 0.001); ADC values correlated negatively with VEGF expression levels (ρ = −0.884, 95% CI -0.971 to −0.736, p < 0.001). When evaluating villous VEGF expression (high, n = 25 vs. medium/low, n = 25), the area under the receiver operator characteristic (ROC) curve (AUC) by dynamic enhancement, ADC values, and their combination was 0.865 (95% CI 0.766–0.964), 0.905 (95% CI 0.821–0.988), and 0.965 (95% CI 0.922–1.000), respectively; after bootstrap internal validation (1000 resamples) the combined model retained an optimism-corrected AUC of 0.956. At Youden-derived cut-offs the corresponding sensitivities were 84% (95% CI 64–95), 88% (95% CI 69–97), and 92% (95% CI 74–99); specificities were 84% (95% CI 64–95), 84% (95% CI 64–95), and 92% (95% CI 74–99). The combined model discriminated significantly better than enhancement amplitude alone (DeLong Z = −2.069, p = 0.039), but not better than ADC alone (Z = −1.687, p = 0.092). Conclusions: MRI dynamic enhancement and DWI are associated with villous VEGF expression in early tubal pregnancy and show promising, internally-derived discrimination. The combined model achieved the highest apparent discrimination, although its advantage over ADC alone did not reach statistical significance in this sample. Whether these indices can guide personalized treatment planning requires prospective, externally validated study.