The aim of this study was to compare the effectiveness of transvaginal power Doppler sonography with spectral Doppler analysis as an aid in preoperatively distinguishing primary ovarian carcinoma and metastatic carcinoma to the ovary (Krukenberg tumors).
Fifty women with ovarian disease were preoperatively examined with transvaginal power Doppler sonography. Six basic parameters were measured, including intratumoral peak systolic velocity, end-diastolic velocity, time-averaged maximum velocity, pulsatility index (PI), resistive index (RI), and velocity index (VeI). Blood flow analyses were detectable in all patients. Twelve patients with metastatic carcinoma to the ovary were classified as group 1; 38 patients with primary ovarian carcinoma were classified as group 2. Comparison of intratumoral blood flow analyses between the two groups was performed.
The PI, RI, and VeI were significantly lower in patients with metastatic carcinoma to the ovary than those with primary ovarian carcinoma (P < .05). There were no significant differences in the peak systolic velocity (P = .871), end-diastolic velocity (P = .508), and time-averaged maximum velocity (P = .850) between the two groups.
Transvaginal power Doppler sonography with spectral Doppler analysis is an effective method in evaluating intratumoral blood flow of Krukenberg tumors. Low impedance (PI, RI, and VeI) might assist us in making differential diagnoses between primary ovarian carcinoma and Krukenberg tumors according to our preliminary results.
It has been more than 100 years since Krukenberg1 originally recognized 6 patients with unusual ovarian tumors as carcinomatodes and thought they were sarcomatous in nature. The term “Krukenberg tumor” has been widely applied to any metastases to the ovary, even though some authors consider that there should be a more strict histologic definition, such as located in the ovaries, with a sarcomatouslike stroma, and composed of signet ring cells with intracellular mucin.2 However, more than one fourth of all ovarian metastases present as probable primary ovarian carcinomas,3 and 5% to 10% of all ovarian malignancies are metastatic carcinomas, which are difficult to discriminate.4
Several reports in the literature have shown the possibility of differential diagnosis from the characteristics of sonography, computed tomography, and magnetic resonance imaging.5–8 The application of transvaginal sonography, color flow imaging, and Doppler waveform analysis allows a closer approach to pelvic hemodynamics. Various blood velocity measurements have been used to show the low impedance of intratumoral circulation in ovarian malignancy.9 Although excellent results in differentiating benign from malignant tumors have been reported by some authors, there are others who have described less encouraging findings.10–13 One previous report analyzed primary ovarian cancer and metastatic tumors to the ovary by transvaginal gray scale and color Doppler sonography.14 It revealed that the presence of a purely solid tumor indicates a higher probability of metastatic carcinoma than primary ovarian cancer. However, with the use of gray scale and color Doppler sonography, it is difficult to differentiate primary ovarian carcinomas from metastatic tumors to the ovary. To the best of our knowledge, none of the previous reports provided us with data from blood flow analysis within Krukenberg tumors. We assessed intratumoral blood flow analysis from 12 Krukenberg tumors using power Doppler sonography with spectral Doppler analysis and made comparisons with primary ovarian carcinoma.
Krukenberg tumors are widely defined as the presence of any metastasis to the ovaries. Among the primary lesions in the gastrointestinal tract, the colon is the most common site in Western countries, and the stomach is the most common site in Asian countries.15,16 The other sites of primary tumors include the breast, pancreas, lung, gallbladder, small intestine, and kidney, as well as melanoma, sarcoma, and carcinoid tumors.17 The clinical incidence is approximately 5% to 10% in the
and 15% to 20% in Asia.18–20 The presence of ovarian
metastasis is universally a poor prognostic sign and commonly occurs during a
woman’s reproductive years. Early, prompt, and aggressive therapy is beneficial
for the patient’s quality of life and symptom relief. Asymmetrically enlarged
bilateral encapsulated masses, variable intratumoral echogenic density, and
ascites are the major characteristics on sonography.
In 1975, Ingersoll and Scully21 first described a patient with ovarian carcinoma, metastatic from the colon, with a complex lesion containing cystic and solid features on combined B-scan and A-mode sonography. Rochester et al5 also described a patient with a Krukenberg tumor (colon origin) that showed homogeneous low-level echoes with the usual gain, and excellent transmission of sound was seen on low-gain and high-frequency (3.5-MHz) sonography. Later, Choi et al15 studied 16 Krukenberg tumors from patients with gastric carcinoma. They found varied echogenicity within tumors on sonography: solid in 8 patients, mixed in 6, and predominantly cystic in 2. Shimizu et al22 further classified Krukenberg tumors in 9 patterns by the tumor wall, solid part, and cystic part. They found that 14 of 15 Krukenberg tumors had clear tumor margins, irregular hyperechoic solid patterns, and “moth-eaten” cyst formations on real-time gray scale sonography.22 In our study, 5 Krukenberg tumors had clear tumor margins, but only 2 patients (40%) had irregular hyperechoic solid parts; 3 patients (60%) had diffuse homogenous solid parts around internal small clear cystic formations. Only 1 patient had a moth-eaten cyst formation; the others had irregular cystic formations.
Transabdominal color Doppler sonography has provided us with a new assessment method in Krukenberg tumors. An abnormal vascular pattern with high-velocity, low-impedance signals within heterogeneous solid masses was the important sonographic characteristic.23 Moreover, Cho et al24 showed that power Doppler sonography was superior to conventional color Doppler sonography in detecting slow blood flow, especially in showing relatively prominent vascular signals along the walls of intramural cysts, which were located in solid masses in 2 patients with Krukenberg tumors from gastric carcinoma. Additionally, 3-dimensional (3D) transvaginal power Doppler imaging better defined the morphologic and vascular characteristics of ovarian lesions.25 The ovarian malignancies were correctly identified by both 2-dimensional and 3D imaging; however, the specificity significantly improved with the addition of 3D power Doppler sonography. To the best of our knowledge, no previous studies reported that 3D power Doppler indices could be useful tools in differentiating primary ovarian cancer and metastatic ovarian tumors. Alcázar26 reported that vascularization, as assessed by 3D power Doppler sonographic indices, was higher in advanced stage and metastatic ovarian cancers than in early-stage ovarian cancer. However, no differences were found in the PI, RI, and PSV in early-stage ovarian cancer, advanced ovarian cancer, and metastatic ovarian cancer.26
In our study, the mean PI and RI in the group with primary ovarian cancer showed no significant differences with previous reports.27,28 However, the mean PI and RI values of Krukenberg tumors were lower than those of primary ovarian tumors (P < .05). We may assume that the vasculature of metastatic ovarian cancer is much different from that of primary ovarian tumors and normal ovarian vessels. Active angiogenesis might be the key factor of these findings. After attaching to ovaries, the metastatic cells activate the quiescent vasculature to produce new blood vessels with wall structures that have little or no smooth muscle support.28 These abnormal vessels, along with arteriovenous shunting in some metastatic tumors, contribute to diminished vascular resistance and, therefore, are reflected by PI and RI values that are much lower than those of primary ovarian tumors. Furthermore, increased permeability of abnormal vessels could cause stasis of blood flow and short shunts, which result in low impedance.In this study, the patients with Krukenberg tumors had a significantly lower VeI (PSV/EDV) than those with primary ovarian tumors. This result might indicate that metastatic tumors could not establish organized intratumoral vasculature, especially when they seeded on other organs or tissues. The unhealthy vasculature reflected the lower VeI. On the contrary, primary ovarian tumors could have been directly and abundantly nourished by ovarian vessels, which resulted in the higher PSV.
In conclusion, power Doppler sonography might be a useful tool in evaluating intratumoral blood flow analysis of ovarian carcinoma before surgery. According to our results, lower impedance (such as a lower PI, RI, and VeI) could be a guideline when making differential diagnoses between primary and metastatic ovarian carcinoma. However, further studies are needed to determine the cutoff values of the PI, RI, and VeI for differentiating primary and metastatic ovarian tumors.