The Use of Unenhanced Doppler Sonography in the Evaluation of Solid Breast Lesions

Transcription

1 Unenhanced Doppler Sonography of Solid Breast Lesions Jose L. del Cura 1 Elena Elizagaray Rosa Zabala Ana Legórburu Domingo Grande del Cura JL, Elizagaray E, Zabala R, Legórburu A, Grande D Received July 8, 2004; accepted after revision September 22, All authors: Servicio de Radiodiagnóstico, Hospital de Basurto, Av. Montevideo 18, Bilbao 48014, Spain. Address correspondence to J. L. del Cura (jdelcura@hbas.osakidetza.net). AJR 2005;184: X/05/ American Roentgen Ray Society The Use of Unenhanced Doppler Sonography in the Evaluation of Solid Breast Lesions OBJECTIVE. The objectives of our study were to investigate differences in Doppler sonography features between benign and malignant breast lesions and between malignant lesions with different prognostic factors and to propose diagnostic criteria for Doppler sonography of breast lesions. SUBJECTS AND METHODS. We performed power and duplex Doppler sonography examinations in 826 breast lesions scheduled for sonographically guided core needle biopsy. Lesion vascularity, pulsatility index (PI), and resistive index (RI) of the vessels detected were analyzed and correlated with histologic results. RESULTS. Color flow was more frequently seen in malignant (237/348 lesions, 68%) than in benign (171/478, 36%) lesions (p < 0.001). However, sensitivity, specificity, and positive and negative predictive values for this sign were low (68%, 64%, 58%, and 73%, respectively). The RI and PI values were significantly higher (p < 0.001) in cancers. Although an overlap in these values between benign and malignant lesions was observed, all but one nodule with an RI of greater than 0.99 (those with null or inverted diastolic flow) or a PI of greater than 4 were malignant. No significant relationship was found between PI, RI, or flow visualization on power Doppler sonography and tumor grade or lymph node involvement in cancers. CONCLUSION. Flow visualization on power Doppler sonography indicates a higher possibility of malignancy but is not useful as the main sign for malignancy. However, any lesion with a vessel that has an RI value greater than 0.99 or a PI value greater than 4 within it must be considered as probably malignant regardless of any other sonography sign present. Doppler findings are not useful to predict tumor grade or lymph node involvement. onography is currently one of the S main diagnostic methods for diagnosing breast diseases. Using this technique has become common practice since the introduction of high-frequency probes; nowadays, it is included in most routine procedures to detect and identify breast lesions. Similarly, sonography is one of the most widely used imaging techniques for guiding breast intervention procedures such as needle biopsy or marking lesions for surgery. Since the article by Stavros et al. [1] was published on 750 solid breast lesions in which the sonographic criteria for differentiating between benign and malignant lesions were defined, the features from grayscale sonography that are used to determine the nature of breast lesions and its significance are widely accepted. However, there is no such consensus when it comes to the use of Doppler sonography to diagnose these lesions. Several studies have used different Doppler sonography criteria both qualitative and quantitative to differentiate between benign and malignant lesions of the breast and to predict various prognostic factors such as axillary infiltration or grade of the tumor [2 11]. These investigations have used both pulsed and color Doppler sonography. The results obtained have not always matched, so the usefulness of Doppler sonography in diagnosing breast cancer is not currently defined. To help define the role of Doppler sonography in the diagnosis of breast lesions, we present the results of a study performed on more than 800 breast nodules. The aim of this study was to determine the usefulness of both pulsed and power Doppler in predicting the malignancy of lesions and determining the prognosis of carcinomas. We used power Doppler sonography instead of color Doppler sonography because of its higher sensitivity for detecting vessels [5] AJR:184, June 2005

2 Unenhanced Doppler Sonography of Solid Breast Lesions Fig year-old woman with fibroadenoma. Power Doppler sonogram shows no color signals. Subjects and Methods Patients Between January 2000 and July 2002, all patients who showed one or more solid masses on sonography and were scheduled to undergo sonographically guided needle core breast biopsy were included in a prospective study. The study entailed performing an examination of the lesion using both gray-scale and Doppler sonography before performing the biopsy. Our radiology service offers to perform breast biopsies on all lesions classified as BI-RADS category 3, 4, or 5 detected on mammography [12] and on all lesions that do not fulfill the criteria of a cyst on sonography. If these lesions are visible on sonography, the biopsy is performed using sonographic guidance. Written informed consent was obtained from all patients to perform the breast biopsy. The study was authorized by our hospital s institutional review committee. A Fig year-old woman with invasive ductal carcinoma. Power Doppler sonogram shows vessel that penetrates lesion and branches. Exploration Technique All the sonography explorations were performed by one experienced radiologist using either an ATL HDI 3500 unit (Philips Medical Systems) with a 5 12-MHz linear transducer or a Sonoline Elegra unit (Siemens AG Medical Solutions) with a 5 13-MHz linear transducer, both of which are equipped to perform power Doppler sonography. The lesions were imaged with gray-scale sonography to evaluate their sonographic characteristics and to ascertain their diameters. Next, power Doppler sonography was performed. The color box was adjusted to include the lesion and a small margin of normal breast tissue. The color sensitivity was adjusted so that only the background color was suppressed and small vessels could be detected. During the exploration, care was taken to apply as little pressure as possible with the probe on the lesion to prevent vessels from collapsing. The exploration with power Doppler sonography was considered positive if at least one vessel was depicted inside the lesion and it showed an arterial flow pattern when explored with pulsed Doppler imaging (Figs. 1 and 2). If intratumoral vascularization was detected, then a study of the vessels detected using pulsed Doppler sonography was performed to obtain spectral waveforms for up to three different vessels. The B Fig year-old woman with fibroadenoma. A, Gray-scale sonogram reveals lobulated heterogeneous nodule. B, Duplex Doppler sonogram of vessels in tumor shows low-resistance pulsatile signal. AJR:184, June

3 resistive index (RI) and pulsatility index (PI) values were calculated for all these waveforms (Fig. 3). The RI value was calculated as the maximum systolic velocity minimum systolic velocity / maximum systolic velocity. The PI value was calculated as the maximum diastolic velocity minimum diastolic velocity / mean velocity. The calculations were done using the equipment s software by placing the cursors on the traces obtained and selecting the best cycle from the Doppler waveform. The highest RI and PI values obtained were considered the RI and PI values for the lesions. All data relative to the characteristics of the lesion obtained by means of gray-scale sonography, pulsed Doppler sonography, and power Doppler sonography were recorded before the results for the biopsies on the lesions were known. Diagnosis A sonographically guided core needle biopsy was performed on all the lesions using a device (Pro-Mag 2.2 L, Manan Medical Products) with disposable needles. For those patients who underwent surgery, the final diagnosis was obtained by means of histologic examination of the surgical specimen. In malignant lesions, histologic grade and stage of the tumor were determined, and the presence of lymph node involvement was assessed. The diagnosis from the core needle biopsy was taken as the final diagnosis for those patients who did not undergo surgery or for those for whom it was not possible to obtain the surgical report. All patients who had a benign finding at core needle biopsy and did not undergo surgery were followed up by means of clinical or radiologic control studies or by telephone calls at least 1 year after the procedure. No cancers developed in the biopsied lesions during that period. For statistical purposes, phyllodes tumors were always considered to be malignant regardless of their histologic features. Statistical Analysis Computer software (SPSS 11.0, Statistical Package for the Social Sciences) for Windows (Microsoft) was used for statistical analysis. The Mann-Whitney U test for independent data was used to compare RI and PI values between benign and malignant lesions and to compare the maximum diameters of lesions with or without detectable vascularization. The RI and PI values in lesions with and without lymph node involvement were analyzed using the Student s t test. A oneway analysis of variance was performed to analyze the differences in the RI and PI values for breast carcinomas with different grades of tumor. The detection of vascularization using power Doppler sonography in benign or malignant lesions, palpable and nonpalpable lesions, and malignant tumors with or without axillary lymph node involvement was analyzed using the chisquare test. Only those lesions for which pathologic data had been gathered as a result of axillary lymphadenectomy were considered to analyze the relationship between the various Doppler parameters and the existence, or not, of lymph node involvement. Logistic regression analysis was performed to determine whether the detection of flow in the tumor was a significant predictor of breast malignancy. For this analysis, the diameter and the palpability of the lesion were controlled, given that both factors have shown significant differences between malignant and benign lesions. In addition to those data, the following statistical performance data were calculated for each of the diagnostic tests: sensitivity, specificity, positive predictive value, and negative predictive value. To calculate these values and their confidence interval, we used a software package (Epidat 2.0, Xunta de Galicia). The findings with a p value of less than 0.05 were considered to be statistically significant. TABLE 1 Final Diagnoses Diagnosis No. of Cases Normal breast tissue 35 Gynecomastia 5 Cyst or fibrocystic breast disease 91 Adenosis 19 Fibroadenoma 261 Ductal or lobular hyperplasia 8 Intraductal papilloma or papillomatosis 10 Fat necrosis 10 Focal fibrosis 8 Lipoma 2 Inflammatory tissue 13 Tubular adenoma 14 Radial scar 2 Invasive lobular carcinoma 37 Ductal carcinoma in situ 4 Invasive ductal carcinoma 239 Tubular carcinoma 2 Papillary carcinoma in situ 1 Invasive papillary carcinoma 10 Medullary carcinoma 1 Mucinous carcinoma 11 Invasive carcinoma (not otherwise 17 specified) Sarcoma or phyllodes tumor 7 Lymphoma 4 Metastatic adenopathy a 12 Breast cancer in men 3 Total 826 a In all cases, metastatic adenopathy corresponded to breast nodules that did not appear as clear lymph nodes on sonography. Results The sonographic examination and core needle biopsy were performed on 732 patients. The number of lesions for each patient varied between one and three, resulting in a total of 826 lesions. The patients ranged in age between 17 and 94 years (mean age, 51 years; SD, 15 years). Of these, 270 patients were sent from our community s breast cancer screening program, which is performed for women between the ages of 50 and 65 years. Another eight patients were male. Of the lesions biopsied, 463 corresponded to palpable tumors and 363 to nonpalpable tumors. The lesions appeared on the sonographic examinations as solid, well-circumscribed nodules (n = 413); lobulated nodules (n = 147); masses with irregular or spiculated edges (n = 150); masses with ill-defined edges (n = 66); nodules on the inside of dilated ducts (n = 12); or intracystic nodules (n = 38). The largest diameter of the lesions varied between 3 and 99 mm (average, 18 mm; SD, 11 mm). Surgical excision was performed on 332 of these lesions. It was not possible to obtain the histologic report of the surgical specimen in 12 of the patients (with a total of 17 lesions) whose biopsies showed malignant tumors and who underwent surgery at an outside hospital. Of those patients who underwent surgery and whose histologic report of the surgical specimen was available, only four with a benign result in the biopsy were found to have a malignant tumor at surgery; these four cases were a phyllodes tumor that was diagnosed as a fibroadenoma in the core needle biopsy; an infiltrating papillary carcinoma that had been diagnosed as papilloma in the biopsy; and two infiltrating ductal carcinomas, one of which had been diagnosed in the core needle biopsy as ductal hyperplasia and the other as focal fibrosis. None of the lesions that was considered to be malignant in the biopsy proved to be benign at surgery. On 48 of the lesions that had been diagnosed as malignant at biopsy, no surgical excision was performed. Chemotherapy or radiation therapy was performed on 32 of these 48 lesions; in the remaining 16 cases, the patients refused to receive any treatment. The final diagnoses of the lesions are shown in Table 1. In total, 478 lesions were benign and 348 were malignant. In 251 cases of the latter group, the results of axillary lymph node dissection were gathered and showed lymph 1790 AJR:184, June 2005

4 Unenhanced Doppler Sonography of Solid Breast Lesions TABLE 2 Comparison of Vascular and Nonvascular Breast Lesions Lesion Characteristic Vascular Nonvascular Total p Nature of lesion < No. (%) of benign lesions 171 (36) 307 (64) 478 No. (%) of malignant lesions 237 (68) 111 (32) 348 Palpable < No. (%) of palpable lesions 304 (66) 159 (34) 463 No. (%) of nonpalpable lesions 104 (29) 259 (71) 363 Maximum diameter (mm) < Average SD Axillary involvement Positive node 85 (71) 35 (29) 120 Negative node 80 (61) 51 (39) 131 TABLE 3 Diagnostic Value of Various Findings on Doppler Sonography as a Sign of Malignancy % (95% Confidence Interval) Doppler Finding Sensitivity Specificity PPV NPV Detection of vascularization 68 (63 73) 64 (60 68) 58 (53 63) 73 (69 78) Resistive index value 1 14 (10 19) 99 (96 100) 97 (82 100) 45 (40 51) Pulsatility index value > 4 10 (7 15) 99 (96 100) 96 (78 100) 44 (39 50) Note. PPV = positive predictive value, NPV = negative predictive value. Resistance Index Benign Malignant A Pulsatility Index Benign node metastasis in 120. Of the 244 carcinomas that were operated on and for which the histologic grade could be ascertained, 35 were grade I; 140, grade II; and 69, grade III. Vessels could be detected on the power Doppler examination in 171 (36%) of the 478 benign lesions and 237 (68%) of the 348 malignant lesions (Table 2). The difference between detecting vascularization in benign and malignant lesions using Doppler sonography was statistically significant (p < 0.001). The difference between detecting vessels in palpable and nonpalpable lesions was also statistically significant (p < 0.001): For the palpable lesions, 66% of the lesions displayed vascularization, whereas 29% of the nonpalpable lesions were also vascularized (Table 2). In addition, the maximum diameter of lesions with detectable vascularization (mean, 22.2 mm; SD, 13.3 mm) was greater than in nonvascular lesions (mean, 13.5 mm; SD, 7.2), and the difference was statistically significant with a p value of less than (Table 2). Nevertheless, analysis by means of logistic regression showed that detection of vascularization was significantly more frequent in malignant lesions regardless of the size of the lesions studied or whether the lesion was palpable or not (p < 0.001). Detection of vascularization on power Doppler sonography as a sign to predict malignancy (Table 3) showed a sensitivity of 68%, a specificity of 64%, a positive predic- Malignant B Fig. 4. Resistance and pulsatility index distribution in benign and malignant lesions. A, Scattergram shows overlap of RI in benign and malignant lesions under level of 1. Only one benign lesion (a radial scar) had RI over B, Scattergram shows that all but one (same previously described radial scar) of benign lesions had PI values lower than 4. All lesions with PI higher than 4, except that one, were malignant. AJR:184, June

5 Fig year-old woman with invasive ductal carcinoma. Duplex Doppler sonogram shows high-resistance pulsatile signal, with inverted diastolic flow, in vessels inside tumor. tive value of 58%, and a negative predictive value of 73%. In palpable lesions, 193 of the 255 malignant lesions and 111 of the 208 benign lesions displayed vascularization. In this group of lesions, detection of vascularization on power Doppler sonography had a sensitivity of 76% and a specificity of 47%. For the nonpalpable lesions, 44 of the 93 malignant lesions and 60 of the 270 benign lesions displayed vascularization, with a sensitivity rating for this group of lesions of 47% and a specificity of 78%. Of the five noninfiltrating carcinomas, vascularization was detected in three. Vascularization was detected in 26 (74%) of the 35 grade I tumors, in 80 (57%) of the 140 grade II tumors, and in 55 (80%) of the 69 grade III tumors. We did not observe significant differences in the detection of vascularization between malignant nodules with axillary lymph node involvement and those without it, detecting vascularization in 85 of the 120 malignant nodules with lymph node involvement and in 80 of the 131 lesions with no axillary involvement (Table 2). In those lesions in which vessels had been detected and for which it had been possible to obtain Doppler spectral waveforms, the RI value (Fig. 4) varied between 0.35 and 1.46 (mean, 0.82; SD, 0.18) for the 237 malignant lesions, whereas for the 171 benign lesions, the RI value varied between 0.26 and 1.51 (mean, 0.65; SD, 0.15). Both the RI and PI values (Fig. 4) were significantly higher in malignant lesions (p < in both). Of the 33 lesions with an RI equal to or greater than 1 (i.e., those with null or reversed diastolic flow), only one (3%), a radial scar, was benign (Fig. 5). Of the 25 lesions with a PI value greater than 4, only the same lesion was found to be benign (4%). The detection of an RI equal to or greater than 1 as a sign of malignancy presented a specificity of 99% and a positive predictive value of 97% both very high, but it was observed in a relatively small number of cases; thus, the sensitivity and negative predictive value were low (14% and 45%, respectively). A PI value of greater than 4 had similar values: sensitivity, 10%; specificity, 99%; positive predictive value, 96%; and negative predicative value, 44% (Table 3). There were no significant differences between the RI and PI values for lesions with axillary lymph node involvement and those without it or between lesions of different histologic grades. Discussion Tumor angiogenesis plays an important role in the growth and extension of malignant neoplasms, including those of the breast [13, 14]. The increase of tumor mass and the formation of metastasis require the formation of new vessels. Malignant tumors secrete angiogenic factors to help the recruitment on new vessels that allow tumoral enlargement. Therefore, it seems reasonable that a technique that allows vascularization assessment, such as Doppler sonography, could be used to differentiate between benign and malignant breast lesions [15] and even to predict the prognosis of the tumors. Until now, various parameters of Doppler sonography have been analyzed to determine their usefulness in the diagnosis of breast lesions. The detection of vascularization in the lesion was the first Doppler feature used to differentiate between benign and malignant tumors, and this sign has shown a significant association with malignancy of lesions [2, 4, 5, 7, 11]. However, higher detection rates of flow in malignant tumors are not necessarily due to tumoral angiogenesis: There are also other breast nodule characteristics that are significantly associated with more frequent detection of flow on Doppler sonography, such as the size of the lesion or whether it is palpable. This has been observed by previous investigators [6, 7, 9] and can be due to the fact that the detection of vessels on the inside of a nodule is easier if the lesion being studied is close to the transducer, as frequently occurs with palpable lesions. Also, it seems to be more likely to detect vessels in a lesion that, because of its larger size, contains more and larger vessels. In fact, in the cases we have studied, malignant tumors have shown a significant larger average size and have been palpable more frequently than benign lesions. Nevertheless, the logistic regression study in our data confirms that, controlling for these factors, the malignant nature of the tumor itself is related to higher detection rates of vessels using power Doppler sonography on breast cancers. Previous reports have shown a marked variability for detecting vessels in a tumor as a sign of malignancy, with sensitivities ranging from 73% to 98%, specificities from 16% to 90%, positive predictive values from 41% to 85%, and negative predictive values from 47% to 99% [2 5, 7, 9, 11, 16, 17]. This variability may be explained by the differences in selecting which lesions to study within each one of the trials, which include different proportions of malignant and palpable lesions. In any case, our results, which to our knowledge correspond to the largest series yet published, are discouraging and suggest little practical usefulness of the detection of flow in breast tumors as a sign of malignancy. Although malignant tumors showed vascularization significantly more often than benign ones, we were unable to detect vessels in 32% of the breast cancers that we studied, whereas 34% of the benign lesions displayed vascularization on the Doppler examination. Even among palpable lesions, for which the sensitivity of this sign is greater, sensitivity reached no more than 76%, which is clearly inadequate to rule out malignancy in any lesion in which no flow is detected. New vessels that originated as the result of tumor angiogenesis have characteristics that differ from those of normal vessels. These 1792 AJR:184, June 2005

6 Unenhanced Doppler Sonography of Solid Breast Lesions vessels are characterized by their lack of muscle layer, atypical branching pattern, and irregularity, and they frequently present stenosis, occlusions, or arteriovenous fistulas [18]. These vascular anomalies produce alterations in the dynamics of the arterial flow that can be evaluated by means of pulsed Doppler sonography. In those nodules in which we were able to obtain flow velocity curves, we found significant differences between the flow patterns for malignant and benign lesions, the former having much higher RI and PI values than the latter. This has also been observed by other authors [3, 11]. We hypothesize that this increase in resistance is related to the existence of occlusions and stenoses in the tumor vessel network structure produced by vascular encasement due to tumor growth. However, the RI and PI values for malignant and benign lesions overlap significantly, which makes these values show little usefulness as a tool to differentiate between both types of lesions. The disappearance or reversal of vascular flow in the diastole, which indicates the presence of very high resistance flow, appears to be closely connected to malignancy in the lesions we studied, given that 32 (97%) of the 33 tumors in which this sign appeared were carcinomas and the remaining lesion corresponded to a radial scar, which is also a lesion to be surgically removed. This sign, to our knowledge, has not been previously described and, when present, the RI is equal to 1 (if the diastole A Fig year-old woman with invasive lobular carcinoma. A, Gray-scale sonogram reveals homogeneous, well-circumscribed, slightly hyperechoic solid mass with slight posterior enhancement. Lesion had appearance of a typically benign lesion. B, However, duplex Doppler sonogram shows high-resistance flow, with inversion of flow in diastole. shows no flow) or greater than 1 (if the flow in the diastole is reversed). Although infrequent it occurred in only 14% of our breast cancer cases, this sign, given its high positive prediction value, could contribute to the characterization of breast lesions on sonography because if it was observed in a lesion, that lesion would be considered as probably malignant despite any other sonographic or mammographic characteristics (Fig. 6). This disappearance or reversal of vascular flow can be explained as an extreme case of the previously described phenomena of vascular occlusion, encasement, and stenosis, typical of malignant vascularity causing a great increase in the peripheral resistance of the vascular bed of some of these tumors. Research has also shown that the detection of flow can be a prognostic factor in the study of breast cancer. In particular, the detection of vessels in a tumor has been connected with the involvement of axillary lymph nodes and with lymphatic vascular invasion. Some authors have discovered significant differences in the detection of vessels between those breast carcinomas with axillary lymphatic invasion and those without it, having found vessels by means of Doppler sonography in 74 91% of tumors with axillary lymph node infiltration, whereas 90% of tumors that displayed no vascularization did not have axillary involvement [6, 10]. Nevertheless, other authors have not found any significant relationship between color Doppler detection of vessels in tumors and lymph node involvement [2, 7]. We found flow in 72% of the breast neoplasms with lymph node involvement a percentage similar to some of those previously described and the proportion of vascularization detected in these tumors was greater than that of tumors that did not have lymph node involvement, which displayed a vascularization rate of 60%. Nevertheless, the differences between both groups were not enough to be statistically significant. Only 62% of the cancers in which vascularization was not detected did not have involvement of lymphatic nodes. Moreover, we have been unable to show any significant differences between the RI and PI values in breast cancers with or without axillary involvement. Therefore, the results of our study indicate that Doppler sonography has no usefulness for predicting axillary lymph node involvement in breast cancer. Tumor grade is another of the prognostic factors that can be associated with the detection of vascularization in a breast carcinoma using Doppler sonography. In a previous study, it was described that the vascularization displayed with color Doppler sonography was greater in grade III breast cancers than in the rest [7]. In our study, the percentage of vessels detected was similar in grade I and III tumors, which both presented vascularization more often than grade two tumors. B AJR:184, June

7 Therefore, just as with other authors [2, 6], we have been unable to establish a relationship between vascularization and tumor grades, and we have been unable to find any significant differences in the RI and PI of tumors with different grades. This study has various limitations. First, not all the patients underwent surgery, which means that the diagnosis for most comes from the core needle biopsy. However, in the cases that were operated on, core needle biopsy showed a nearly perfect correlation with the results obtained from surgery, with only four false-negatives of 332 lesions. In addition to this, the follow-up period at least 1 year for benign lesions that were not operated on allowed us to confidently rule out the existence of undetected malignant lesions. Thus, it seems that the use of the core needle biopsy results as a definitive diagnosis method in patients who did not undergo surgery will not significantly influence the outcome of this study. Although our study has shown that the detection of a null or reversed diastolic flow can be strongly associated with malignancy in the tumors in which it is detected, the proportion of carcinomas in which this sign has been found is relatively small, suggesting that further studies would be required to determine the usefulness of this sign. Recently, various authors have proposed the use of contrast-enhanced Doppler sonography to study breast lesions [8, 16, 17]. The use of contrast material amplifies the detection of vessels in lesions with Doppler sonography, and it can increase the sensitivity of vessel detection in the tumor as a sign of malignancy. The use of contrast-enhanced power Doppler sonography has managed to detect vessels in up to 95% of malignant tumors [17]. A high sensitivity would allow this sign to be used to avoid biopsies of lesions in which vascularization had not been detected by means of contrast-enhanced power Doppler sonography. However, in our opinion, the use of IV contrast agents would turn a cheap, quick, and innocuous examination into a more expensive, time-consuming, and slightly invasive one, with no substantial gain in diagnostic information if compared with core needle biopsy, only slightly more aggressive. In conclusion, Doppler sonography is, by itself, of little use when it comes to evaluating solid breast lesions. However, when it is used in conjunction with conventional sonography examinations, it can help provide a more accurate characterization of certain lesions. The detection of vessels on the inside of a breast nodule is significantly linked with malignancy. The detection of a diastole with null or reversed flow on spectral Doppler imaging is a sign, albeit infrequent, of very high positive predictive value; thus, when that sign is present, the lesion should be classified as probably malignant. On the other hand, Doppler sonography is useless to predict the prognosis of patients with breast carcinomas. References 1. Stavros AT, Thickmann D. Rapp CL, Dennis MA, Parker SH, Sisney GA, Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. Radiology 1995; 196: Cosgrove DO, Kedar RP, Bambar JC, et al. Breast diseases: color Doppler US in differential diagnosis. Radiology 1993;189: Peters-Engl C, Medl M, Leodolter S. The use of color-coded and special Doppler ultrasound in the differentiation of benign and malignant breast lesions. Br J Cancer 1995;71: Raza S, Baum JK. Solid breast lesions: evaluation with power Doppler US. Radiology 1997;203: Kook SH, Park HW, Lee YR, Lee YU, Pae WK, Park YL. Evaluation of solid breast lesions with power Doppler sonography. J Clin Ultrasound 1999;27: Mehta TS, Raza S. Power Doppler sonography of breast cancer: does vascularity correlate with node status or lymphatic vascular invasion? AJR 1999;173: Holcombe C, Pugh N, Lyons K, Douglas-Jones A, Mansel RE, Horgan K. Blood flow in breast cancer and fibroadenoma estimated by colour Doppler ultrasonography. Br J Surg 1995;82: Yang WT, Metrewelli C, Lam PKW, Chang J. Benign and malignant breast masses and axillary nodes: evaluation with echo-enhanced color power Doppler US. Radiology 2001;220: Birdwell RL, Ikeda DM, Jeffery SS, Jeffery RB. Preliminary experience with power Doppler imaging of solid breast masses. AJR 1997; 169: Kubek KA, Chan L, Frazier TG. Color Doppler flow as an indicator of nodal metastasis in solid breast masses. J Ultrasound Med 1996; 15: McNicholas MMJ, Mercer PM, Miller JC, Mc- Dermott EWN, O Higgins NJ, McErlean DP. Color Doppler sonography in the evaluation of palpable breast masses. AJR 1993;161: American College of Radiology. Breast Imaging Reporting and Data System (BI RADS), 3rd ed. Reston, VA: American College of Radiology, Gasparini G, Harris AL. Clinical importance of the determination of tumor angiogenesis in breast carcinoma: much more than a new prognostic tool. J Clin Oncol 1995;13: Weidner N, Semple U, Welch WR, Folkman J. Tumor angiogenesis and metastasis: correlation in invasive breast carcinoma. N Engl J Med 1991; 324: Weind KL, Maier CF, Rutt BK, Moussa M. Invasive carcinomas and fibroadenomas of the breast: comparison of microvessel distributions implications for imaging modalities. Radiology 1998; 208: Sturhman M, Aronius R, Schietzel M. Tumor vascularity of breast lesions: potentials and limits of contrast-enhanced Doppler sonography. AJR 2000;175: Moon WK, Im JG, Noh DY, Han MC. Nonpalpable breast lesions: evaluation with power Doppler US and a microbubble contrast agent initial experience. Radiology 2000;217: Less JR, Skalak TC, Sevick EM, Jain RK. Microvascular architecture in a mammary carcinoma: branching patterns and vessel dimensions. Cancer Res 1991;51: AJR:184, June 2005