This retrospective study received approval from the Ethics Committee of People’s Hospital of Yuxi City (approval number: 2025kmykdx6f060). From December 2017 to October 2024, there were a total of 34,140 patients who underwent cesarean section in our hospital, among whom 19,912 had vaginal deliveries and 14,228 had cesarean sections.

The diagnostic criteria for POVT [1, 8] included postpartum fever, lack of response to simple anti-infective treatment, thrombosis confirmed by contrast-enhanced CT, clinical improvement following combined anti-infective and anti-thrombotic therapy, and recovery confirmed by follow-up at 3 months.

Postpartum women who underwent a CT scan from the top of the diaphragm to the lower edge of the pubic symphysis within the 42-day postpartum period were included in this study [11].

The exclusion criteria were as follows:

(1) Patients with ovarian vein tumors.

(2) Patients who have undergone internal fixation surgery around the ovarian vein.

A 64-slice CT scanner (GE Lightspeed VCT, GE Healthcare, Waukesha, WI, USA) was used for 16 POVT cases and 36 non-POVT cases. Scan parameters included 120 kV, 250 mA, a 0.6-second rotation time, 40 mm detector width, 0.984:1 pitch, and a 512 $\times$ 512 matrix. For the remaining 8 POVT cases and 6 non-POVT cases, third-generation CT scanners (SOMATOM Force, Siemens Healthcare, Shanghai, China) were utilized with 120 kV, 240 mA, 0.5-second rotation time, 96 mm detector width, and a 512 $\times$ 512 matrix. Scanning was performed from the top of the diaphragm to the lower edge of the pubic symphysis. After the non-contrast scans, 90–100 mL of Ioversol (catalogue No. H20067896; 320 mgI/mL, Jiangsu Hengrui Pharmaceuticals, Lianyungang, Jiangsu, China), followed by 40 mL of normal saline, was injected through an elbow vein indwelling needle at a rate of 4 mL/s.

All CT images were analyzed using a GE workstation (ADW 4.7, GE Healthcare, Waukesha, WI, USA). The slice thickness and inter-slice interval were both set at 0.625 mm.

2.3.2 Non-Contrast CT Semi-Quantitative Parameters

A senior radiologist with 21 years of experience in gynecological imaging measured the ovarian vein width and the CT values of both the ovarian vein and inferior vena cava. Measurements were taken three times on different slices, and the average was used (Figs. 1A,2A,3A,C,D). The CT attenuation difference was defined as the CT value of the ovarian vein (on the side with higher density and greater width) minus the CT value of the inferior vena cava. The ratio was calculated by dividing the ovarian vein CT value (same side) by the inferior vena cava CT value.

Fig. 1.

A 24-year-old woman with right ovarian vein thrombophlebitis. The red arrow indicates the ovarian vein, the yellow arrow indicates the inferior vena cava, and the green arrow indicates the periovarian vein exudate. (A) Non-contrast axial CT image. The right OV is approximately 1.4 cm wide (CT value = 52.00 HU, red arrow). The inferior vena cava shows a CT value of 32.40 HU (yellow arrow), with visible surrounding exudation (green arrow). (B) Contrast-enhanced sagittal CT image. A filling defect is observed in the right ovarian vein, consistent with POVT, and the vessel wall appears uniformly thickened (red arrow). (C) Contrast-enhanced axial CT image. The right ovarian vein shows a clear filling defect, and the vessel wall is uniformly thickened (red arrow). (D) Three-dimensional vein reconstruction. The inferior vena cava is marked (yellow arrow), with the right OV shown as a thin red arrow and the left OV as a thick red arrow.

Fig. 2.

A 17-year-old woman with left ovarian vein thrombophlebitis. The red color (red circle, red arrow and red cuboid) indicates the ovarian vein, the yellow color (yellow circle and yellow arrow) indicates the inferior vena cava. (A) Non-contrast axial CT image. The right OV measures approximately 0.91 cm in width (CT value = 46.10 HU, red arrow). The CT value of the inferior vena cava is 32.00 HU (yellow arrow). (B) Non-contrast coronal CT image showing the left OV (red arrow). (C) Contrast-enhanced axial CT image. Thrombosis in the left OV with uniform wall thickening (red arrow). (D) Contrast-enhanced coronal CT image. Left POVT with uniformly thickened wall (red arrow).

Fig. 3.

The primary and secondary CT findings of ovarian vein thrombophlebitis. The red arrow indicates the ovarian vein, the yellow arrow indicates the inferior vena cava, and the green arrow indicates the periovarian vein exudate. (A) Non-contrast axial CT image. The right ovarian vein is approximately 2.92 cm wide (CT value = 63.50 HU, red arrow). The CT value of the inferior vena cava is 38.40 HU (yellow arrow), with adjacent exudation (green arrow). (B) Non-contrast axial CT image showing pelvic effusion (thick red arrow). (C,D) A 21-year-old woman with right ovarian vein thrombophlebitis and inferior vena cava thrombosi: (C) Non-contrast axial CT image. The right ovarian vein is about 1.6 cm wide (CT value = 59.30 HU, red arrow). The inferior vena cava shows a CT value of 71.80 HU (yellow arrow), and surrounding exudation is noted (green arrow). (D) Non-contrast axial CT image.

Statistical analyses were conducted using SPSS software, version 27.0 (IBM Corp, Armonk, NY, USA). The mode of delivery was assessed using the chi-square test, with a statistical significance set at p 0.05. CT plain scan data, including age, ovarian vein width, CT attenuation difference, OV/inferior vena cava (IVC) attenuation ratio, and D-dimer levels, were analyzed. The Shapiro-Wilk test was applied to assess the distribution of each variable. If the data followed a normal distribution, the T-test was used; if not, the Mann-Whitney U test was applied. Statistical significance was defined as p 0.05.

Twenty-four POVT cases were identified (23 right-sided, 1 left-sided), with an age range of 26–43 years (mean $\pm$ SD: 30.88 $\pm$ 7.05). Twenty-four POVT cases met the inclusion criteria, with no excluded cases. A control group comprised 42 postpartum women without POVT (aged 19–38 years, mean $\pm$ SD: 29.19 $\pm$ 4.10). Among the 42 non-POVT patients, all reported significant abdominal pain. Underlying causes included abdominal trauma (n = 2), pancreatitis (n = 4), pancreatic tumor (n = 1), suspected appendicitis (n = 3), suspected ureteral calculi on ultrasonography (n = 2), and suspected intestinal obstruction in the remaining cases. The normal axillary temperature of a healthy person is 36.5 $\pm$ 0.4 °C, and an axillary temperature of 37 °C is the cut-off point between normal and fever [12]. All 24 POVT patients presented with fever, showing axillary temperatures from 37.1 °C to 38.8 °C (median: 38.00 °C; interquartile range: 37.45 °C to 38.3 °C). In comparison, the 42 non-POVT patients also had fever, with temperatures ranging from 36.8 °C to 39.2 °C (median: 37.15 °C; interquartile range: 36.9 °C to 37.65 °C). A statistically significant difference was found between the two groups (p = 0.02). Thirteen patients showed elevated white blood cell counts. Right lower quadrant pain was reported in 13 cases, while 5 patients experienced abdominal distension and pain accompanied by vomiting, indicative of intestinal obstruction. There was no statistically significant difference in mode of delivery between the POVT and non-POVT groups (p $>$ 0.99, Table 1), nor was there a significant difference in age (p = 0.22).

In the POVT group, the D-dimer was determined by latex immunoturbidimetry (STAGO, RMA, Paris, France). The mean postpartum D-dimer level was 7.66 $\pm$ 4.94 µg/mL. Using 3.24 µg/mL as the reference threshold [13], 21 patients had elevated D-dimer levels, while 3 were within the normal range. In the non-POVT group, the mean level was 1.94 $\pm$ 0.91 µg/mL, with 4 patients showing elevated values. The difference in D-dimer levels between the two groups was statistically significant (Z = –5.49, p 0.01).

The primary CT findings of POVT included filling defects within the ovarian vein lumen along with thickening of the vessel wall (Figs. 1,2,3). Secondary signs involved pelvic effusion in 13 cases and periovarian venous exudation with increased fat density in 19 cases (Figs. 1A,C,3A,C). One patient, who developed extensive thrombosis extending to both lower extremity veins, the inferior vena cava, and pulmonary arteries, died from postpartum hemorrhagic shock. The remaining 23 patients fully recovered.

Compared with the non-POVT group, the POVT group showed significantly greater ovarian vein width on the thrombus side, higher CT attenuation of the thrombus-side ovarian vein, as well as increased CT attenuation difference and ratios (p = 0.001, Figs. 1,2,3). Using 0.90 cm as the cut-off value for ovarian vein width, the sensitivity reached approximately 91.70%, with a specificity of 95.20% (Figs. 1A,2A,3A,C,D). With 46.00 HU as the threshold for ovarian vein CT value on the thrombus side, both sensitivity and specificity were 100% (Figs. 1A,2A,3A). For a CT attenuation difference cut-off of 12.50 HU, the sensitivity was 95.30%, and specificity remained 100%. Similarly, a ratio cut-off of 1.1 yielded a sensitivity of 95.30% and specificity of 100% (Figs. 1A,2A,3A).

In one POVT case involving extensive inferior vena cava thrombosis, the CT value of the inferior vena cava exceeded that of the ovarian vein, resulting in a negative CT attenuation difference of –12.50 HU and a ratio of 0.83 (Table 1; Fig. 3C,D).

Among the 42 non-POVT cases, the mean diameter of the right ovarian vein was 0.65 $\pm$ 0.22 cm, while the left ovarian vein measured 0.66 $\pm$ 0.19 cm. The difference between the two sides was not statistically significant (p = 0.72).

POVT was first reported by Austin [14] in 1956 and remains a rare and often underrecognized condition. In this study, the incidence of POVT was 0.06% (24 out of 34,140 cases), with a rate of 0.05% (10/19,912) following vaginal delivery and 0.10% (14/14,228) after cesarean section. These findings align with previously published data, where POVT incidence ranged from 0.002% to 0.05% overall [13]. Other studies suggest that POVT occurs in approximately 0.05%–0.18% of vaginal deliveries and 1%–2% of cesarean deliveries [15]. In our cohort, all 24 patients experienced fever, with axillary temperatures between 37.1 °C and 38.8 °C. Fever is one of the most common clinical features of POVT [16, 17, 18], consistent with earlier reports showing its presence in up to 80% of patients [16].

This study has several limitations. First, as a single-center retrospective analysis, the use of data from only one institution may introduce selection bias. Second, the small sample size—limited to 24 confirmed POVT cases—reduces statistical power and increases the likelihood of false negatives. Third, although CT-based parameters showed high sensitivity and specificity overall, variations between individual cases were observed. Fourth, although the semi-quantitative CT parameters demonstrated high diagnostic accuracy, the thresholds (e.g., CT value $\ge$46.00 HU, OV width $\ge$0.90 cm) were derived from a limited dataset and may be subject to overfitting. These cut-off values might not perform equally well in broader clinical settings with greater heterogeneity. External validation using larger, multi-center datasets is essential to confirm the robustness and reproducibility of the proposed diagnostic criteria. Fifth, while experienced radiologists conducted the image interpretation, the study did not formally assess interobserver variability. Variations in measurements such as ovarian vein width or CT attenuation values across readers could impact diagnostic consistency.