Can cross-sectional imaging replace diagnostic venography in pelvic venous disorder (PeVD)?

Conception and design: MS, OA, AA, SA, KA Analysis and interpretation: MS, OA, AA, SA, KA Data collection: MS, OA, AA, SA, KA Writing the article: MS, OA, AA, SA, KA Critical revision of the article: MS, OA, AA, SA, KA Final approval of the article: MS, OA, AA, SA, KA Statistical analysis: MS, OA, AA, SA, KA Obtained funding: MS, OA, AA, SA, KA Overall responsibility: MS

The research proposal was reviewed and approved by the “Committee of Medical Ethics” of the Faculty of Medicine-Assiut University. A female patient presented by either one or more of the following: symptomatic varicose veins along the back of the thigh, buttock, and vulvoperineal varicosities, chronic pelvic pain of unknown cause at least 6 months, congestive dysmenorrhea, dyspareunia, postcoital ache, and urinary symptoms, left flank or abdominal pain that was worsened by standing, sitting, or walking, hematuria, and recurrence of symptomatic lower limb varicosities after superficial venous ablations by average 1 month. Also, those patients had at least one imaging modality using either CTV or MRV prior to venography. Patients with varicose veins along the course of great and small saphenous veins or with infrainguinal sources of refluxing lower limbs varicose veins (deep system reflux or perforators) on DUS were excluded. Also, patients with an absolute contraindication to CTV, MRV, and diagnostic venography or a history of lower limbs deep venous thrombosis (DVT) or prior hysterectomy and oophorectomy were excluded. Medical records and notes were checked for clinical evaluation via a history premenopausal female who has chronic pelvic pain associated with menses, and coitus, which increased by standing for long periods and decreased by lying flat. Other causes of chronic pelvic pain of other etiologies (such as urological, gynecological, etc causes) were excluded. Gynecological diseases were excluded, namely endometriosis, adenomyosis, and chronic pelvic inflammatory disease. Then, all patients had a duplex examination. If duplex suggests a diagnosis of PCS or inconclusive infrainguinal reflux source, patients were referred for further evaluation using a cross-sectional study to confirm the diagnosis and exclude obstructive pathologies such as May-Thurner and Nutcracker syndromes. Then conventional venography was performed to confirm the diagnosis and to evaluate the pelvic venous dynamics and, at some times, perform embolization procedures. DUS records were reviewed to exclude the source of reflux in the lower limbs. The competence of various veins, such as the sapheno-femoral junction and the sapheno-popliteal junction, were evaluated, as well as deep venous system patency and reflux. Pelvic varicocele, crossing myometrial veins, and pelvic venous plexus diameters were also assessed using abdominal and transvaginal approaches. Ovarian veins reflux or diameter was not assessed by duplex, only changes in waveform by colored duplex during Valsalva, which is a surrogate for reflux. The duplex scans were performed using different transducers with a comprehensive protocol of transabdominal/transvaginal ultrasound. TR-MRV was done for the group of patients (group A) with pelvic venous incompetence in whom no infrainguinal source was found for varices on duplex ( Fig 1 ). Magnetic resonance imaging scan was conducted with a 1.5 T imaging unit (Magnetom Sempra, Siemens) with the use of a body phased-array coil. To evaluate the pelvic veins and structures, axial, coronal, sagittal T2-weighted turbo spin-echo images and axial T1-weighted fast low-angle shot images were applied. Contrast agent (0.2 mL/kg of Magnevist; Bayer-Schering) was injected intravenously at a rate of 2 ml/s with saline flush in due course. After a bit, the four phases (arterial, late arterial, venous, and late venous) were carried out in the coronal plane when patients breathed shallowly. Furthermore, the transit time of the contrast medium was set utilizing a bolus tracking technique to accomplish maximum venous contrast signal. The MRV scan was post-processed utilizing maximum intensity projection. Additionally, to achieve the abdominopelvic post-contrast axial three-dimensional (3D) fat-suppressed T1-weighted sequence. Retention of contrast medium was >20 seconds and included filling of vulvovaginal varices. Fig 1 Images of a 39-year-old multiparous female with suspected pelvic congestion syndrome (PCS). A , Magnetic resonance venography with time-resolved imaging (TR-MRV) image showing dilated left ovarian vein (OV) of 8.5 mm diameter and ipsilateral pelvic parauterine veins with at least one measuring 6 mm crossing the midline. B , Selective catheter-directed venography confirming the incompetence and dilatation of left OV with congestion in the pelvic venous plexus, retention of contrast medium >20 seconds, and filling of vulvovaginal varices. Images of a 39-year-old multiparous female with suspected pelvic congestion syndrome (PCS). A , Magnetic resonance venography with time-resolved imaging (TR-MRV) image showing dilated left ovarian vein (OV) of 8.5 mm diameter and ipsilateral pelvic parauterine veins with at least one measuring 6 mm crossing the midline. B , Selective catheter-directed venography confirming the incompetence and dilatation of left OV with congestion in the pelvic venous plexus, retention of contrast medium >20 seconds, and filling of vulvovaginal varices. Also, multi-detector CTV with 3D reconstruction was performed for a group of patients (group B). CTV scan was executed using a 32-32-multi-detector CT (Sensation 16; Siemens Medical Solutions). First, we performed an intravenous injection via the peripheral vein of 2 mL/kg (maximum, 150 mL) of nonionic Ultravist 370 mg I/mL contrast agent (Schering AG) at a rate of 2.5 mL/s, utilizing a power injector. After that, we determined the scan delay time to achieve the maximum venous contrast and the scan range from the upper renal pole to the end of the feet. To generate the 3D reconstruction, the individual volume data were uploaded into the 3D program and fused with axial images for noise reduction and interactive volume rendering. For the sake of post-processing of the scans, multiple planar reconstructions, volume rending, and curved plane reconstruction of the venous phase images were applied ( Fig 2 ). Fig 2 Images of a 30-year-old multiparous female with suspected pelvic congestion syndrome (PCS). A , Computed tomographic venography (CTV) with three-dimensional (3D) reconstruction image showing left-sided May-Thurner Syndrome with incompetent internal iliac vein (IIV) and multiple dilated pelvic varicosities. B , Selective catheter-directed venography confirming the incompetence of left IIV with ipsilateral dilated and tortious parauterine pelvic veins measuring 1.5 cm. Images of a 30-year-old multiparous female with suspected pelvic congestion syndrome (PCS). A , Computed tomographic venography (CTV) with three-dimensional (3D) reconstruction image showing left-sided May-Thurner Syndrome with incompetent internal iliac vein (IIV) and multiple dilated pelvic varicosities. B , Selective catheter-directed venography confirming the incompetence of left IIV with ipsilateral dilated and tortious parauterine pelvic veins measuring 1.5 cm. Diagnostic venography was performed for all patients in the current study under sterile conditions and regional anesthesia using fluoroscopic control in an angiography suite. In a reverse-Trendelenburg position, unilateral or bilateral femoral vein access was established percutaneously with the Seldinger technique. After insertion of the 6-French vascular sheath, catheter-directed selective venography of the inferior vena cava (IVC), internal iliac veins (IIVs), and ovarian veins (OVs) was carried out utilizing a 6-Fr × 65 cm Cobra catheter (Cook) over a 0.035-inch guidewire. Multiplanar projection was used to assess iliac vein compression or Nutcracker syndrome. In evaluating suspected Nutcracker or May-Thurner syndrome, attention is paid to the left renal vein and iliac veins to detect the refluxing OV and IIV, respectively. To enhance and detect the incompetent OV, IIV, and pelvic venous plexus, 20 ml diluted nonionic Ultravist 300 mg I/mL contrast agent (Schering AG) was manually injected through the catheter with Valsalva maneuver. All multi-detector CTV, TR-MRV, and diagnostic venography images were interpreted retrospectively by two independent personnel; one was an experienced vascular surgeon (M.S), whereas the second was an experienced interventional radiologist (A.A.) to increase reliability and decrease interobserver variability interpretation bias related to interobserver disparities. Diagnostic venography was used as the standard of reference. The diagnostic appearance of pelvic varicoceles by DUS was tortuous and dilated veins that were greater than 5 mm in diameter around the ovary and uterus with initial flow accentuation during the Valsalva maneuver. 10 , 11 , 12 , Diagnostic criteria for PCS on cross-sectional modalities depend on ovarian (OVR) and iliac vein reflux (IIVR). On MRV, grade I venous reflux is defined by early filling of left OV and/or left para-uterine veins, whereas grade II venous reflux includes grade, I reflux plus right OVR or IIVR (left and/or right). 12 On CTV/MRV, grade III reflux is defined as finding an OV diameter of >8 mm, and grade IV reflux is marked by identifying four ipsilateral pelvic veins with at least one measuring >4 mm. 13 Images were assessed for the presence of polycystic ovarian configuration, organic masses, enlarged lymph nodes, or any lesions that may cause chronic pelvic pain or vascular obstruction and secondary PCS. No significant lesions were found. Demonstration of the venous anatomy on multi-detector CTV and TR-MRV was assessed as either inadequate (impossible to definitively determine a treatment plan), sufficient (intermediate image quality but sufficient for treatment planning), or adequate anatomic visualization. We applied the classification method according to Kaupilla, 14 defining the venographic appearance of the ovarian and iliac veins as well as the para uterine venous plexus as normal (small, straight, similar in caliber, and easily visualized veins), moderate congestion (vein variable in caliber, tortuous, and challenging to see separately, diameter between 0.5 and 1.0 cm), or severe congestion (wide veins, significant variation in caliber, markedly tortuous, diameter greater than 1.0 cm). Inadequate image referred to the absence of any lesions that may cause chronic pelvic pain or vascular obstruction and secondary PCS, failure of visualization of OVR and IIVR and para uterine venous plexus, whereas sufficient referred to the absence of any lesions that may have caused chronic pelvic pain or vascular obstruction and secondary PCS, visualization of either OVR or IIVR in addition to visualization para uterine venous plexus. Lastly, adequate image was identified as the absence of any lesions that may cause chronic pelvic pain or vascular obstruction and secondary PCS successful visualization of both OVR and IIVR in addition to visualization of the para uterine venous plexus. The assessment of OVs refluxes their diameters and drainage location to the renal vein or IVC. The maximum diameters of both OVs were measured in the axial plane. The presence of one or more of the following findings on diagnostic venography is highly suggestive 15 ; OV diameter >6 mm, or contrast medium retention for >20 seconds, or congestion in the pelvic venous plexus, or opacification of the ipsilateral (or contralateral) IIV, or filling of vulvovaginal and thigh varices, or free reflux down the left OV and into the parametrial veins upon left renal vein injection in reverse Trendelenburg position.

The present study was conducted in the vascular surgery department and the diagnostic and interventional radiology department, Assiut University Hospital, Egypt during the period from March 2021 through September 2022. This retrospective observational cohort study of the maintained medical records was conducted on a selected 120 patients whose final diagnosis was PCS as established by the diagnostic venography. Patients underwent a duplex scan of the saphenofemoral junction and the lower limbs’ deep and superficial veins system. Afterward, all selected patients underwent CTV or MRV examination before the diagnostic venography. The Institutional Review Board approved the study protocol developed following the Declaration of Helsinki. Medical data were gathered for demographic data and risk factors for developing PCS.

This retrospective observational cohort study of the maintained medical records was conducted on a selected 120 patients whose final diagnosis was PCS. The patients were assigned into two groups: the 60 consecutive patients who presented to surgery and the 60 consecutive patients who presented to radiology. Group A, which is defined as the radiology department group, included 60 patients who underwent TR-MRV. In contrast, group B, which was the vascular department group, included 60 patients who underwent multi-detector CTV-3D. Then, all patients had their diagnostic venography within at least 1 month as a standard of reference. There were no complications attributable to the preceding investigations. Because there are well-known complications to diagnostic venography and the burden on our operative theatre with a long waiting list, which delays patient management, this raised the need for replacing this modality. DUS was reviewed in all patients, and positive findings in DUS with abdominal and transvaginal screening are summarized in Table I . Seventy-two of all patients (60.0%) had pelvic varicoceles with a diameter >5.5 mm, and suspected left-sided May-Thurner syndrome was seen in 16 cases (13.3%). Of all the patients examined, 20.0% had crossing veins in the myometrium, whereas incompetent perforators and SFJ due to refluxing epigastric and/or pudendal tributaries were noted in 16.7% and 33.3% of patients, respectively. Additionally, associated diseases were depicted by abdominal and transvaginal ultrasound: 13.3% had endometriosis, and 6.7% had fibroid. Table I Duplex ultrasound (DUS) with abdominal and transvaginal findings No. (%) Pelvic varicoceles with diameter >5.5 mm 72 (60.0) Crossing veins in the myometrium 24 (20.0) Change in the waveform with Valsalva in OV 64 (53.3) Suspected left May-Thurner syndrome (velocity ratio ≥2.5) 16 (13.3) Incompetent SFJ due to refluxing epigastric and/or pudendal tributaries 40 (33.3) Incompetent perforators 20 (16.7) Endometriosis 16 (13.3) Polycystic ovary 16 (13.3) Fibroid 8 (6.7) OV , Ovarian vein; SFJ , sapheno-femoral junction. Duplex ultrasound (DUS) with abdominal and transvaginal findings OV , Ovarian vein; SFJ , sapheno-femoral junction. In comparing group A with group B, patients who had CTV were more likely to have higher body mass index (27.4% vs 26.2%; P  = .04), hypertension, be multiparous (80% vs 36.7%; P  < .001), and more likely to present with venous claudication (50% vs 16.7%; P  < .001). Patients who had MRV had a trend to be younger (32.3% vs 34.4%; P  = .07), diabetic, and present with dysmenorrhea (50% vs 26.7%; P  = .008) and dyspareunia (46.7% vs 30%; P  = .06). CTV was more likely to show internal iliac vein dilation (60% vs 40%; P  = .028), internal iliac vein reflux (73.3% vs 60%; P  = .121), and adequate venous anatomy (93.3% vs 86.7%; P  = .223) ( Table II ). Table II Comparison between magnetic resonance venography with time-resolved imaging ( TR-MRV ) and three-dimensional computed tomographic venography ( CTV-3D ) for evaluation of pelvic congestion syndrome (PCS) in the current study CTV-3D (n = 60; 50%) TR-MRV (n = 60; 50%) P value Age, years 34.4 ± 7.1 32.3± 5.3 .07 BMI, kg/m 2 27.4 ± 3.4 26.2± 3.2 .04 DM 4 (6.7) 9 (15.0) .141 HTN 10 (16.7) 7 (11.7) .432 Multiparity 48 (80.0) 22 (36.7) .00001 Chronic pelvic pain 60 (100.0) 60 (100.0) Dyspareunia 18 (30.0) 28 (46.7) .060 Dysmenorrhea 16 (26.7) 30 (50.0) .008 Irritable urinary bladder 18 (30.0) 10 (16.7) .084 Atypical varices 60 (100) 60 (100.0) Recurrence after ablation 1 month 8 (13.3) 6 (10.0) .569 Venous claudication 30 (50.0) 10 (16.7) .0001 Dilatation of pelvic veins  OVs 32 (53.3) 40 (66.7) .136  IIVs 36 (60.0) 24 (40.0) .028  Pelvic plexus 24 (40.0) 28 (46.7) .461 Incompetent pelvic veins  OVs 32 (53.3) 36 (60.0) .461  IIVs 44 (73.3) 36 (60.0) .121  Pelvic plexus 32 (53.3) 40 (66.7) .136 Venous anatomy  Sufficient 4 (6.7) 8 (13.3) .223  Adequate 56 (93.3) 52 (86.7) Venous congestion  Normal 28 (46.7) 20 (33.3) .135  Moderate 8 (13.3) 16 (26.7)  Severe 24 (40.0) 24 (40.0) BMI , Body mass index; DM , diabetes mellitus; HTN , hypertension; IIV , internal iliac vein; OV , ovarian vein. Data are presented as number (%) or mean ± standard deviation. Comparison between magnetic resonance venography with time-resolved imaging ( TR-MRV ) and three-dimensional computed tomographic venography ( CTV-3D ) for evaluation of pelvic congestion syndrome (PCS) in the current study BMI , Body mass index; DM , diabetes mellitus; HTN , hypertension; IIV , internal iliac vein; OV , ovarian vein. Data are presented as number (%) or mean ± standard deviation. In comparing MRV with venography, MRV was less likely to visualize dilated ovarian veins (66.7% vs 86.7; P  = .009) and pelvic plexus (46.7% vs 66.7%; P  = .027). Venography was more likely to detect ovarian vein incompetence (73.3% vs 60%; P  = .121) and pelvic plexus incompetence (80% vs 66.7%; P  = .003). Venography was superior in demonstrating moderate and severe venous congestion with a P value of .034 ( Table III ). In comparing CTV with venography, CTV was more likely to detect IIV dilations (60% vs 33.3%; P  = .003) and IIVR (73.3% vs 40%; P  < .001). Venography was more likely to visualize OV dilations (80% vs 53.3%; P  = .001) and pelvic plexus dilations (60% vs 40%; P  = .028). Also, it was superior in demonstrating moderate and severe venous congestion with a P value of .002 ( Table III ). Table III Comparison between magnetic resonance venography with time-resolved imaging ( TR-MRV ) and three-dimensional computed tomographic venography ( CTV-3D ) vs venography for evaluation of pelvic congestion syndrome (PCS) in the current study TR-MRV group CTV-3D group TR-MRV (n = 60; 50%) Venography (n = 60; 50%) P value CTV-3D (n = 60; 50%) Venography (n = 60; 50%) P value Dilatation of pelvic veins  OVs 40 (66.7) 52 (86.7) .009∗ 32 (53.3) 48 (80.0) .001∗  IIVs 24 (40.0) 12 (20.0) .016∗ 36 (60.0) 20 (33.3) .003∗  Pelvic plexus 28 (46.7) 40 (66.7) .027∗ 24 (40.0) 36 (60.0) .028∗ Incompetent pelvic veins  OVs 36 (60.0) 44 (73.3) .121 32 (53.3) 24 (40.0) .143  IIVs 36 (60.0) 16 (26.7) .002∗ 44 (73.3) 24 (40.0) .0002∗  Pelvic plexus 40 (66.7) 24 (80.0) .003∗ 32 (53.3) 40 (66.7) .136 Venous anatomy  Inadequate 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)  Sufficient 8 (13.3) 0 (0.0) 4 (6.7) 0 (0.0)  Adequate 52 (86.7) 60 (100) 56 (93.3) 60 (100) Venous congestion  Normal 20 (33.3) 8 (13.3) .034∗ 28 (46.7) 12 (20.0) .002∗  Moderate 16 (26.7) 20 (33.3) 8 (13.3) 20 (33.3)  Severe 24 (40.0) 32 (53.4) 24 (40.0) 28 (46.7) IIV , Internal iliac vein; OV , ovarian vein. Data are presented as number (%). Comparison between magnetic resonance venography with time-resolved imaging ( TR-MRV ) and three-dimensional computed tomographic venography ( CTV-3D ) vs venography for evaluation of pelvic congestion syndrome (PCS) in the current study IIV , Internal iliac vein; OV , ovarian vein. Data are presented as number (%). Additionally, we did not notice any findings related to Nutcracker syndrome and inferior vena cava anomalies on both CTV and diagnostic venography modalities. Intravenous ultrasound (IVUS) was not used in our study as it is costly and unavailable in our institution. Also, the pressure gradient was not evaluated, as Nutcracker syndrome is a rare condition in our locality, and we used visual evaluation only. Other pathology detected by CT during examination as 16 cases (13.3%) had endometriosis, 16 patients (13.3%) had polycystic ovaries, and eight patients (6.7%) had fibroid. Utilizing diagnostic venography as the gold standard, we identified the sensitivity, specificity, and diagnostic accuracy of multi-detector CTV with 3D reconstruction and TR-MRV, as presented in Tables IV and V . CTV imaging had sensitivity, specificity, and diagnostic accuracy of 50%, 33%, and 47% for the detection of incompetent OV; 83%, 33%, and 53% for the detection of incompetent IIV; and 50%, 40%, and 47% for detection of incompetent pelvic plexus veins, respectively. Cohen’s k-values indicated poor agreement for identifying the refluxing OV and pelvic plexus veins (0.003, 0.002) and a slight agreement for identifying the refluxing IIV (0.2). Table IV Sensitivity, specificity, and diagnostic accuracy of multi-detector computed tomographic venography ( CTV ) with three-dimensional ( 3D ) reconstruction compared with diagnostic venography Indices of CTV-3D Incompetent veins OVs IIVs Pelvic plexus Sensitivity, % 50.0 83.3 50.0 Specificity, % 33.3 33.3 40.0 Accuracy, % 46.7 53.3 46.7 PPV, % 33.3 45.4 62.5 NPV, % 49.9 74.9 28.5 IIV , Internal iliac vein; NPV , negative predictive value; OV , ovarian vein; PPV , positive predictive value. Table V Sensitivity, specificity, and diagnostic accuracy of magnetic resonance venography with time-resolved imaging ( TR-MRV ) compared with diagnostic venography Indices of TR-MRV Incompetent veins OVs IIVs Pelvic plexus Sensitivity, % 72.7 75.0 66.7 Specificity, % 25.0 45.5 33.3 Accuracy, % 60.0 53.3 60.0 PPV, % 72.6 33.3 80 NPV, % 25.0 83.3 31.2 IIV, Internal iliac vein; NPV, negative predictive value; OV, ovarian vein; PPV, positive predictive value. Sensitivity, specificity, and diagnostic accuracy of multi-detector computed tomographic venography ( CTV ) with three-dimensional ( 3D ) reconstruction compared with diagnostic venography IIV , Internal iliac vein; NPV , negative predictive value; OV , ovarian vein; PPV , positive predictive value. Sensitivity, specificity, and diagnostic accuracy of magnetic resonance venography with time-resolved imaging ( TR-MRV ) compared with diagnostic venography IIV, Internal iliac vein; NPV, negative predictive value; OV, ovarian vein; PPV, positive predictive value. On the other hand, TR-MRV had sensitivity, specificity, and diagnostic accuracy of 73%, 25%, and 60% for the detection of incompetent OV; 75%, 46%, and 53% for the detection of incompetent IIV; and 67%, 33%, and 60% for detection of incompetent pelvic plexus veins, respectively. Also, Cohen’s k-values expressed slight agreement for pinpointing the refluxing OV and IIV (0.03, 0.05) and almost perfect agreement for pinpointing the refluxing pelvic plexus veins (0.89).

Diagnosis of PCS is a matter of challenge for health care workers due to its indefinite clinical symptoms. Misdiagnosis may put the patients in a frustrated situation because of the consequences that may affect their quality of life. Until today, no proof exists that any available imaging modalities are superior for diagnosing PCS. Diagnostic venography is a particular concern due to its complications that may be related to the use of intravenous access, ionizing radiation, and iodinated contrast material. In our study, the patients were of childbearing age with a mean age of 34.4 ± 7.1 years. About 86.7% of patients were multipara, and the mean number of deliveries is two (range, 0-7). This may be related to the fact that there is a correlation between PCS and increased parity and ovarian activity during the reproductive age. 18 , 19 , 20 All patients in the current study had pelvic pain associated with varicose veins in the vulva, perineal, buttock, and back of the thigh. In 2011, guidelines endorsed using DUS scanning as an initial diagnostic test, but more advanced imaging was recommended if necessary. DUS offers benefits such as dynamic imaging and availability. In the current study, 20% of patients had recurrent varicosities after surgery, and incompetent SFJ and perforators were noted in about 33% and 17% of patients, respectively. This is consistent with other studies that found the SFJ to be one of the areas most often affected by recurrent reflux after surgery. 21 , 22 , 23 , 24 However, CTV is of limited utility in evaluating the reflux status of pelvic veins. 24 , 25 No statistically significant difference between CTV/MRV and diagnostic venography was found in the present study regarding the dilatation and incompetence of pelvic veins. Concerning the venous anatomy, we achieved adequate visualization in 93.3% with CTV and 86.7% with MRV compared with diagnostic venography, which showed 100%. Additionally, moderate to severe congestion was noted in 53.3% of CTV and 66.7% of MRV. In comparison, diagnostic venography recorded it at 80% and 86.7% in both CTV and MRV, respectively, with no statistically significant difference. Similar results of Asciutto et al 26 that compared MRV with diagnostic venography showed adequate venous anatomy in 91% vs 96% and moderate to severe congestion in 57% vs 74% of women, respectively. Unfortunately, to date, not enough data are available about the sensitivity and diagnostic accuracy of CTV in the evaluation of pelvic venous incompetence, as it does not have a common place in the imaging protocol. Ignacio et al 27 concluded that CTV has a greater sensitivity in detecting lower pelvic varicosities than DUS, but it has a lower sensitivity than MRV and diagnostic venography for the diagnosis of PCS. In agreement with this study, our results revealed that the sensitivities of CTV were 50%, 83.3%, and 50% for the OV, IIV, and pelvic venous plexus, respectively. In contrast, the sensitivities of TR-MRV were 72.7% for the OV, 75% for the IIV, and 66.7% for the pelvic venous plexus. So, despite TR-MRV being highly sensitive in all pelvic vein incompetence, CTV with 3D reconstruction has a comparable result and equal accuracy with TR-MRV to detect IIV incompetence only. This may be related to the fact that the time-resolved technique improves the ability of MRV to visualize blood flow dynamics and soft tissue contrast in addition to multiplanar features of MRV that enable us to recognize associated obstructive diseases if compared with CTV. Chennur et al 28 demonstrated pelvic vein incompetence using TR-MRV in 75% of cases (n = 6/8) confirmed by diagnostic venography. Yang et al 12 found no significant difference between TR-MRV and conventional venography in grading OVR with TR-MRV sensitivity, specificity, and diagnostic accuracy of 66.7% to 75%, 100%, and 78.9% to 84.2%, respectively. Dick et al 29 also found that 3D TR-MRA with contrast kinetics (TRICKS) identified reflux better than conventional T2 imaging. TRICKS MRV has better temporal resolution and conspicuity to differentiate between grades I, II, and III reflux. Meneses et al 30 conducted a small study of nine patients prospectively comparing MRV with diagnostic venography and demonstrated that MRV with phase-contrast velocity mapping correctly diagnosed 100% of women with PCS. Correspondingly, Ruehm et al 31 affirmed that MRV had a high sensitivity of 94% for varicose changes in patients with PCS compared with diagnostic venography. Furthermore, we described eight cases of left-sided May-Thurner syndrome using CTV and MRV that were verified by diagnostic venography. However, two cases were diagnosed by diagnostic venography in our population. This finding accords with Chung et al 32 who demonstrated a high correlation between CTV and diagnostic venography in the assessment of iliac venous compression. In studies by Liu et al 33 and Oguzkurt et al, 34 CTV was superior to other imaging modalities in diagnosing May-Thurner syndrome and discriminating between non-thrombotic and thrombotic pathology. In harmony with our data, MRV has been recommended in several studies to diagnose iliac venous compression syndrome. 35 , 36 Nevertheless, McDermott et al 37 found that a single MRV image may not be enough to evaluate iliac venous compression, indicating that another imaging may be needed for further evaluation. They explained that this might be related to several circumstances, such as volume status or patient positioning. Our results revealed no cases of Nutcracker syndrome using cross-sectional imaging and diagnostic venography. Kim et al 38 assessed multiple criteria of Nutcracker syndrome with CTV, including narrowing of the left renal vein at the aortomesenteric portion, the beak sign, the left renal vein diameter ratio of ≥5.0, and the angle between the SMA and aorta of <41° (sensitivity, 66.7%-100%). Similar accuracies of these criteria were achieved with MRV. 48 Therefore, these data assume that CTV with 3D and TR-MRV are valuable options for the morphologic evaluation of pelvic structures and exclusion of the nonvascular causes of pelvic pain in patients with suspected PCS. Generally, the lower sensitivity of these imaging modalities in our study may be related to supine patient positioning that can mask the early reflux in some cases, unlike venography, which allows table tilt and Valsalva maneuver. However, till now, there is no standardized approach to diagnose PCS as the venous flow dynamics in diagnostic venography may be affected by the pressure effect of direct injection, resulting in false-positive results. Our study revealed slight agreement between the TR-MRV and diagnostic venography for pinpointing the refluxing OV and IIV and almost perfect agreement for pinpointing the refluxing pelvic plexus veins. This agrees with Asciutto et al, 26 who found moderate agreement between MRV and diagnostic venography on venous congestion grade and visualization of venous anatomy. On the other hand, a poor agreement between the CTV and diagnostic venography was found in the current study for visualization of the refluxing OV and pelvic plexus veins and slight agreement for the refluxing IIV. In line with our data, the UIP consensus document for the diagnosis and treatment of pelvic congestion advocated for using MRV as a helpful alternative to diagnose pelvic varices and obstructive venous syndromes. 23 Many institutions make patients shift to MRV when there are positive findings with DUS, 39 and MRV can be used as the first imaging for the diagnostic algorithm of PCS. 26 , 27 Recently, the trend in diagnosing PCS is to use both DUS and MRV, which has the advantage over CTV, as it does not need radiation in this generally young population. Hence, CTV is less commonly utilized compared with DUS and MRV. However, CTV is still a valuable alternative when there is an absolute contraindication of the MRV. The results of this study should be interpreted in the context of its limitations. The present study included a limited sample size, thus is at the risk of type II error. The results may be subject to selection bias because this study is retrospective. The outcomes were reported from a single tertiary referral center; thus, the outcomes may not be generalizable to other less experienced centers. The lack of IVUS is a significant limitation in assessing compression syndromes in our study. Also, lack of treatment is considered a significant limitation because uncertainty regarding the patients’ outcomes when certain imaging modalities are not utilized is also a considerable limitation, as the lack of treatment can affect their results.

MRV has a better diagnostic value than CTV, as the former provides sufficient information on hemodynamic changes in pelvic veins when compared with diagnostic venography. However, CTV may offer better visualization of the venous anatomy and diagnosis of other pelvic pathologies and venous compressive syndromes. Although venography is the gold standard for the diagnosis of PCS, cross-sectional imaging is still an important noninvasive method to exclude obstructive etiology and planning of the treatment of PCS.

The medical data were recorded on a sheet form. All statistical tests were performed using SPSS version 24.0 (IBM). Continuous data were expressed in mean and standard deviation, whereas categorical data were expressed in count and percentage. Visualizing venous anatomy and identifying incompetence grades are described as percentages for each imaging modality. Agreement between the noninvasive modalities (CTV or MRV) and diagnostic venography in detecting the refluxing pelvic veins was assessed by Cohen k-values that are interpreted as suggested by Landis. 16 Sensitivity, specificity, and diagnostic accuracy are also reported as percentages. Categorical variables were compared using the χ 2 test. A P value < .05 was considered statistically significant.

None.