INDEX
The Role of Color Doppler Ultrasonography In The Characterization of Ovarian Malignancy

Tayfun Gungor, M.D., Emel Ebru Parlakyiit, M.D.,
Munire Erman Akar, M.D., Oya Gokmen, M.D.,
* Department of Gynecology, Zekai Tahir Burak Maternity Hospital


Corresponding Author
Dr. Tayfun Gungor
Adress for Correspondance:
Akay Cad. 21/10 Bakanl>klar,
Ankara, TURKEY
Fax: 9 312 312 49 31
Total page numbers: 19
Footnote:
Abbreviations used: RI, resistance index.


Key Words: Ovarian malignancy, Doppler ultrasonography, adnexial masses.

SUMMARY
Objective: The aim of the study was to evaluate the effectiveness of doppler flow sonography for the discrimination between benign and malignant ovarian masses and to find out the best RI cut off value.
Material and Methods: In a cross-sectional study 194 patients with adnexial masses were examined using doppler sonography to measure the resistance index of tumor associated blood flow profiles preoperatively.
Results: 74 had malignant tumors, 120 had benign tumors. RI values were ranging from 0,27 to 0,87 and from 0,35 to 1,0 in malignant and benign tumors respectively (p<0,001).
Conclusion: Doppler flow measurements are effective in the differentiation of benign and malignant adnexial masses. Discriminatory zone for the RI is found to be between 0,5 and 0,6, but there is still overlapping between malignant and benign cases and this can be reduced by using tumor markers and two dimensional morphological assesment.

INTRODUCTION
In contrast to the advances in medicine over the last several decades, ovarian malignancies continue to be one of the most important causes of death in gynecologic malignancies. The vast majority of these patients are first identified at stage 3 or stage 4 with widespread metastatic disease. Apart from a reduction in mortality, other possible benefits of early detection is the improvement in the quality of life and the decrease in the cost of treatment. Several diagnostic methods have been introduced to discriminate between benign and malignant ovarian tumors. Ca 125 is used as a marker for nonmucinous ovarian adenocarcinomas., but the positive and negative predictive values are so low that it is mainly used for the postoperative follow-up of patients undergoing chemotherapy.
Two dimensional transabdominal and transvaginal ultrasonography are also accepted tools in clinical decision making but their validity is also restricted because of their limited sensitivity and specifity.
High diastolic flow or low resistance is a pathognomonic feature of vasculature with neoplasms. Neovascular vessels lack of vascular intima form multiple arteriovenous shunts resulting in an increased diastolic flow detectable in the doppler frequency wave form.
More recently, several investigators have been able to distinguish benign from malignant lesions by using color doppler flow mapping and pulsed doppler.
However neovascularity is not specific to malignancy it may also be seen in benign tumors with high proliferative or inflammatory potential. Morphologic assessment of ovarian masses with transvaginal and transabdominal sonography is also used for the detection of malignant tumors however the predictive values are unsatisfactory because of frequent inability to distinguish between malignant and benign tumors with similar morphologic characteristics.
The aim of this study was to evaluate the effectiveness of color doppler sonography in the diagnosis of ovarian malignancy and to find out an optimal cut off value for RI.

MATERIAL AND METHODS
194 pre-post menopausal women with pelvic masses diagnosed by sonography and histopathologic verification formed our study group. Patients were scanned between January 1994 and December 1997 using a Toshiba 140A scanner with a 3.75 or 6 mhz transducer Patients were scanned both transabdominally and transvaginally as necessary for the best technical visualisation of the adnexial mass.
The color doppler sonography was performed after appropriate evaluation of the lesion morphology. Several measurements were taken from each mass from the wall septum or papillation if present and from a solid side or echophic core. Color signals were visualized and the sample volume was placed over the colored vessel obtaining blood flow velocity waveforms. The resistance index (RI) was calculated as systolic peak minus diastolic through divided by systolic peak. The lowest RI from each case was used in the analysis. Pathologic reports were obtained from laparotomy or laparoscopic biopsy. For all the best cut off value for the RI was determined to identify the malignant lesions.
The sensitivity, specifity and + and - p values were determined for different cut off values of color flow assesment for the prediction of malignancy.
Student t-test and chi square tests are used for the statistical analysis.

RESULTS
194 patients were identified sonographically as having adnexial masses. Pathologically 74 had malignant and 120 had benign tumors. The mean age of the patients for the benign and malignant disease were 42.6 (ranging from 22 to 65) and 53.2 (ranging from 18 to 81) respectively. Mean RI for the benign and malignant lesions were 0.76ア1.19 (ranging from 0.35 to 1.0) and 0.45ア0.13 (ranging from 0.22 to 0.87) respectively. Rl was significantly low for the malignant cases (p<0.001). 14 ovarian masses were not included in this study because no flow could be obtained despite numerous attempts. No malignancies were encountered in this group. We described a lesion as having no flow after 20 to 3O minutes of scanning. These masses consisted of 5 cysts with simple epithelia, 4 endometriomas, 2 teratomas, 2 serous and I mucinous cyst adenoma.
The differentiation of cases according to histopathological diagnosis and their mean values are given at table 1.
There were 6 malignant cases with RI <0.30 (1 fibroma, 1 thecoma, 3 serous cyst adenocarcinoma, 1 dysgerminoma).
Accuracy rate was highest %82.5 for the RI cut off value 0.5.

DISCUSSION
Early detection of ovarian cancer remains a challenge. Gynecological examination in combination with ultrasound findings are the basic and widely available procedures of evaluating adnexial masses (1,2).High frequency transvaginal sonography impruves the ability to detect malignant ovarian tumors over that of the transabdominal route. The vaginal approach produces greater image resolution than the abdominal approach(3,4), thus allowing detailed morphologic assesment of ovarian masses. However, the predictive values are unsatisfactory, because of the frequent inability to distinguish between malignant and benign tumors that have similar morphologic characteristics.
An alternative approach has been introduced by serum Ca125 measurements(5,6). The positive and negative predictive values of this marker for ovarian malignancy are generally low(7),so it is mainly used in the postoperative follow-up of patients undergoing chemotherapy. With the addition of menaposal status to these parameters, a malignancy index concept is introduced.
Recently color doppler with pulsed doppler waveform analysis has been used as a tool to identify neovascularity in malignant masses .
Folkman et al demonstrated that tumor angiogenesis factors are essential for the promotion of neovascularization in malignant tumors(8). These vessels have abnormal morphology with arteriovenous anastomoses and vascular walls deficient in muscular elements, which is reflected by low resistance to flow or high diastolic flow.

Transvaginal doppler wave form analysis provides high sensitivity and specifity and is superior to the other methods for preoperative evaluation of ovarian masses. It can accurately discriminate between malignant and non malignant ovarian tumors using a simple measurement of RI in the newly formed intratumoral blood vessels. Moreover , because early development of neovascularity may precede tumor growth screening for ovarian malignancy with transvaginal doppler waveform analysis may detect early ovarian neoplasms before sonography.
The cut off points of the RI in the best differentiation of malignant from benign lesions were 0.5-0.6 in our data.( For RI=0.5 sensitivity is %70.3, specifity is %90 For RI=0.6 sensitivity is %89.2, specifity is %78.2)
A pelvic mass can represent a number of different benign and malignant conditions. The traditional strategy for the establishment of a final diagnosis has been to perform an exploratory laparotomy. But insufficient primary surgery is a poor prognostic factor for ovarian cancer.
Two dimensional sonography and assesment of Ca125 serum levels are well established indices for the prediction of malignancy in adnexial masses. These variables showed a sensitivity of up to 82%,a specifity of up to 93% and positive predictive values between 31-75% in several large cases with the prevelance of malignancy varying between 15-43 %. There is even at least %I 8 false negative and %7 false positive cases in selected patients.
Jacobs et al introduced a malignancy index based on the menopausal status, ultrasound findings and serum levels of Ca125 (13). It has sensitivity of 80% and specifity of %92, and PPV of 83%. The advantage of this MI was its easy application at less specialized departments of gynecology.
Recent reports using doppler flow evaluation promises higher validity compared with two dimensional sonography and USG or Ca 125 levels. Schneider et al studied 55 patients with adnexial masses(14) A resistance index cut off less than 0.8 showed the highest sensitivity %93.8, specifity of 56.4%, however there were 46% false positive cases. Scheneider et al also defined negative value as the most important feature of a test that evaluates a pelvic mass. In every case identified as benign, diagnosis should be accurate especially when surgical intervention may be abandoned. A NPV of 92% for cut off value 0.6 is reasonably high and can be highered with other conventional methods.
Doppler results showed no advantage compared with established techniques. Applying a cut off value of 0.4, Kurjak et al found that 96.4% of all cancers (n=56) were recognized and only one out of 624 benign tumors was wrongly diagnosed (11).
Bromley et al studied 33 post menopausal patients using pulsed doppler alone with a RI limit of 0.6 (15). The sensitivity of predicting malignancy was 66% with a specifity of 81 %.
Several authors using resistance index have indicated that the demarcation between benign and malignant lesions should be 0.6; others have suggested it should be 0.4 and some others 0.8(8,9,10,11,12). Rl cut off levels between 0.5-0.6 have the highest accuracy rate .
It appears that each laboratory has devised its own index based on its own dataset.
The presence of early stage cancers make discrimination between benign and malignant lesions more difficult. Stage 2 tumor could be easily differentiated in our study. This limits the effectiveness of doppler ultrasound as a screening method but it can still be used with high false positive results.
In this study we tried to find out the most reliable cut off value in determining malignancy . Also we tried to show the relationship between the grade or stage of the tumor and the tumoral vascular flow. Menopausal status was not considered in our study and this increased the false positivity. Although we tried to find out an accurate RI cut off value for the discrimination of malignancy, there were still overlapping of benign and malignant cases in our studied cut off levels(8,9).
The cut off points of the RI in the best differentiation of malignant from benign lesions were 0.5-0.6 in our data. (For RI=0.5 sensitivity is %70.3, specifity is %90 For RI=0.6 sensitivity is %89.2, specifity is %78.2)
Doppler monographic evaluation of resistance indexes in the vessels of adnexial masses is a useful method for the discrimination of malignancy But tumor markers, two dimensional ultrasonography and clinical findings must be considered together for the initial intervention for surgery.

REFERENCES
I. Moyze J.W., Rochester D, Sider 1, Shorck K, Krause P. Sonography of ovarian turnors: Predictability of tumor type. AJR 1983;141:985-991.
2. Andolf E, Jorgensen C.A prospective comparison of clinical, ultrasonical and operative findings of the female pelvis. J. Ultrasound Med 1988;7:617- 620.
3. Weiner Z., Thaler I., Beck D., Rottem S. Deutsch M., Brandes J.M. Differentiating malignant from benign ovarian tumors with transvaginal color: flow imaging. Obstet. Gynecol. 1992; 79: 159-162.
4. Tepper R. Zalel Y., Altaras M., Ben-Baruch G., Beyth Y. Transvaginal color doppler ultrasound in the assessment of invasive cervical carcinoma. 1996;60:26-29.
5. Patsner B., Mann W.J. The value of preoperative serum Ca125 levels in patients with a pelvic mass. Am. J. Obstet. Gynecol 1988; 159: 873-876.
6. Malkasian G.D., Knapp R.C., Lavin P.T. Preoperative evaluation of serum Ca125 levels in premenopausal and postmenopausal patients with pelvic masses. Discrimination of benign from malignant disease. Am. J. Obstet. Gynecol. 1988; 159:341-346.
7. Di-Xia C., Schwartz P.E., Xinguo L., Zhan Y. Evaluation of Ca 125 levels in differentiating malignant from benign tumors in patients with pelvic masses. Obstet. Gynecol. 1988;72:23-7.
8. Tekay A., Jouppila P. Validity of pulsatility and resistance indiexes in clasification of adnexial tumors with transvaginal color doppler ultrasound. Ultrasound Obstet. Gynecol. 1992; 2:338-44.
9. Hamper U.M., Sheth S., Abbas F.M., Rosenshein N.B., Aronson D., Kurman R.J. Transvaginal color doppler sonography of adnexial masses. Differences in blood flow impedance in malignant lesions. AJR 1993;160:2225-8.
10. Timor Tritsch I.E., Lerner J.P., Monteguedo A., Santos R. Transvaginal ultrasonographic characterization of ovarian masses by means of color flow directed Doppler measurements and a morphologic scoring system. Am. J. Obstet. Gynecol. 1993 ;168:909-13.
11. Kujak A., Predanic M. New scoring system for prediction of ovarian malignancy based on tansvaginal color Doppler sonography. J. Ultrasound Med. 1992; 1 1:631-8.
12. Kawai M. Kano T., Kikkawa F., Maeada O., Oguchi H., Tomoda Y. Transvaginal Doppler ultrasound with color flow imaging in the diagnosis of ovarian cancer. Obstet. Gynecol. 1992; 79: 163-7.
13. Jacobs I.J., Rivera H., Oram D.H.,-Base R.C. Differential diagnosis of ovarian cancer with tumor markers Ca 125, Ca 15-3 and TAG 72-3. Br. J. Obstet. Gynecol. 1993; 100: 1 120-1124.
14. Schneider Viola L., Schnelder A., Kathryn L.R., Kenneth D.H. Comparison of Doppler with two dimensional sonography and Ca 125 for prediction of malignancy of pelvic masses. 1993;81:983-8.
15. Bromley B., Goodman H.,, Beryl R.B. Comparison between monographic morphology and doppler waveform for the diagnosis of ovarian malignancy. 1994;83:434-7.

Table 1: Malignant masses (74 patients, 74 masses)

Histopathology

Cases

Mean RI

RI range

Serous Cystadenocarcinoma

28

0.47

0.29-0.71

Mucinous Cystadenocarcinoma

4

0.58

0.29-0.87

Endometrial Carcinoma

6

0.52

0.47-0.67

Clear cell Carcinoma

2

0.43

0.41-0.45

Mixed Epithelial Carcinoma

4

0.51

0.32-0.71

Indifferentiated Epithelial Carcinoma

6

0.45

0.33-0.55

Mixed M_llerian Tumor

2

0.44

0.40-0.48

Immature Teratoma

2

0.36

0.22-0.44

Dysgerminoma

6

0.4

0.33-0.55

Granulosa cell Tumor

14

0.36

0.22-0.44

 

Table 2: Bening masses (120 patients, 120 masses)

Histopathology

Cases

Mean RI

RI range

Fibroma

10

0.55

0.36-0.73

Thecoma

4

0.52

0.35-0.69

Mature Cystic Teratoma

14

0.81

0.58-1.0

Endometrioma

14

0.66

0.43-1.0

Serous Cystadenoma

24

0.86

0.63-1.0

Mucinous Cystadenoma

18

0.85

0.58-1.0

Follicular Cyst

14

0.86

0.70-1.0

Corpus Luteum Cyst

12

0.63

0.38-0.93

Tuboovarian Abscess

8

0.70

0.56-1.0

Paraovarian Abscess

2

0.96

0.90-0.98

 

Sensitivity, specifity, PPV, NPV, accuracy rate and p values for predicting malignancy at cut off values 0.3-0.7 are shown at table 3.

Cut off Value

Sensitivity

Specifity

PPV

NPV

Accuracy Rate

P value

0.3

8.1

100

100

63.8

64.9

>0.05

0.35

18.9

100

100

66.7

69.1

<0.001

0.40

35.1

95

81.2

70.4

72.2

<0.001

0.45

56.8

91.7

80.8

77.5

78.4

<0.001

0.50

70.3

90

81.3

83.1

82.5

<0.001

0.55

75.7

85

75.7

85

81.5

<0.001

0.60

89.2

78.2

71.7

92.2

82.5

<0.001

0.65

89.2

75

68.8

91.8

80.4

<0.001

0.70

91.9

68.3

64.2

93.2

77.3

<0.001

 

The relationship between stages and mean doppler RI according to different types of tumors are shown at table 4.

Histopathology

Stagel

(n--20)

Stage2

(n-- 1 2)

Stage3

(n7-37)

Stage4

(n=4)

Serous cyst adenocarcinoma

1

5

20

2

Mucinous cyst adenocarcinoma

-

2

-

2

Endometrioid adenocarcinoma

-

6

-

-

Clear cell carcinoma

-

-

2

-

Mixed epithelial carcinoma

-

-

4

-

Indifferentiated carcinoma

1

-

5

-

Immature teratoma

-

-

2

-

Dysgerminoma

2

-

2

-

Granulosa cell tumor

14

-

-

-

Mixed M_llerian tumor

2

-

-

-

Mean RI

0,39±0,07

0,55±0,18

0,47±0,12

0,35±0,08

 

Mean RI for stage 2 was significantly different from stage1 and stage 4. The relationship between grades and mean doppler RI according to different types of tumors are shown at table 5.

Histopathology

Grade I

N(I 5)

Grade 2

N(40)

Grade 3

N(52)

Undifferentiated

N(3)

Serous cyst adenocarcinoma

10

18

30

-

Mucinous cyst adenocarcinoma

3

12

13

-

Endometrioid adenocarcinoma

-

1

2

-

Clear cell carcinoma

-

3

2

-

Undifferentiated carcinoma

-

-

-

3

Dysgerminoma

-

-

-

-

Endodermal sinus tumor

-

1

-

-

Immature teratoma

1

-

-

-

Malignant lymphoma

-

1

-

-

Granulosa cell tumor

1

2

1

-

Epidermoid carcinoma

-

-

1

-

Malignant mixed m_llerian tumor

-

-

1

-

Metastatic disease

-

2

2

-

Mean RI

0,39±0,05

0,45±0,11

0,49±0,13

0,41±0,12



Nagoya City University
Medical School