Organ-based Tube Current Modulation: Are Women s Breasts Positioned in the Reduced-Dose Zone? 1

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights. Original Research n Thoracic Imaging Stephen Taylor, MD Diana E. Litmanovich, MD Maryam Shahrzad, MD Alexander A. Bankier, MD, PhD Pierre Alain Gevenois, MD, PhD Denis Tack, MD, PhD Organ-based Tube Current Modulation: Are Women s Breasts Positioned in the Reduced-Dose Zone? 1 Purpose: Materials and Methods: To retrospectively determine the potential of organ-based tube current modulation (OBTCM) to reduce the radiation dose delivered to breast tissue by computed tomography (CT) by determining breast angular position in relation to the zones of decreased versus increased radiation. The authors obtained institutional review board approval for this study and patients written informed consent. In two academic centers (center A: Beth Israel Deaconess Medical Center, Boston, Mass; and center B: Hôpital André Vésale, Montignies-le-Tilleul, Belgium), data were collected from clinical thoracic CT examinations performed in 498 women (mean age, 60 years; age range, 18 95 years) in the supine position and 34 women (mean age, 53 years; age range, 18 84 years) in the prone position. One radiologist in each center determined breast tissue location and measured its ier and outer boundaries with respect to the isocenter of the CT examination. The percentages of women with breast tissue within and those with breast tissue outside the zone of decreased radiation delivered by OBTCM were determined. The location of breast tissue was correlated with patient age and with sagittal and coronal diameters of the thorax by using the Student t test, Fisher exact test, and Pearson correlation. Results: Conclusion: None of the women lying in the supine position had the entirety of the breast tissue located within the reduceddose zone. Breast tissue was located in the increased-dose zone in 99% of women lying supine and in 82% of women lying prone. The breast angular position of almost all women was higher than the angular limit of the reduced versus the increased dose in OBTCM. No woman, regardless of supine or prone position, had all breast tissue within the reduced-dose zone. 1 From the Department of Radiology, Hôpital André Vésale, Montignies-le-Tilleul, Belgium (S.T.); Department of Radiology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Mass (D.E.L., M.S., A.A.B.); Department of Radiology, Hôpital Erasme, Brussels, Belgium (P.A.G.); and Department of Radiology, Epicura Clinique Louis Caty, Rue Louis Caty 136, B-7331 Baudour, Belgium (D.T.). Received March 22, 2014; revision requested May 7; revision received May 22; accepted June 3; final version accepted June 18. Address correspondence to D.T. (email: denis.tack@skynet.be). q RSNA, 2014 q RSNA, 2014 260 radiology.rsna.org n Radiology: Volume 274: Number 1 January 2015

Organ-based tube current modulation (OBTCM) is a recently introduced technology designed to reduce radiation delivery at computed tomography (CT) to ventrally located radiosensitive organs (1). This technology is based on a current reduction when the x-ray tube rotates over the anterior third of the body circumference, where the breasts, thyroid, and eyes are located, and, to maintain image quality, a compensatory current increase over the two lateral and posterior thirds of the body circumference (2 4). When OBTCM is used, the default x-y tube current modulation is disabled. In the chest, it is assumed that female breast tissue is within the zone of reduced radiation (the anterior third of the body circumference), and dose reductions to the breasts have been reported in phantoms (5,6). However, the breast position in phantoms, typically centered anteriorly, does not reflect the real breast position in women. Lungren Advances in Knowledge In our retrospective review of chest CT examinations in 498 women lying in the supine position and 34 women lying in the prone position, we found that no woman had the entirety of her breast tissue within the reduceddose zone corresponding to the anterior 120 of the tube rotation. Data also showed that nearly all women (.99%) who underwent scaing in the supine position had at least a part of their breast tissue within the increased-dose zone corresponding to the remaining 240 of the tube rotation. Although small but significant differences in angle measurements of breast tissue limits and in chest diameters reflecting body habitus existed between the two medical centers located on different continents, the proportions of women with breast tissue exposed to an increased CT dose were not different. et al (5) showed that the average angle between breasts was larger than the 120 of potential reduced-dose anterior arc, suggesting that at least a part of the breasts is lying within or immediately next to the increased-dose zone. Matsubara et al (6) took into account the x-ray fan beam and reported that, in angular positions greater than 120 to 130 of the anterior arc, OBTCM, as compared with fixed tube current, was associated with increased skin dose. The purpose of our study, therefore, was to retrospectively determine the angular position of women s breast tissue in comparison with the decreased versus increased radiation dose zones at chest CT examinations. Because OBTCM can be oriented according to prone or supine patient position, and because breasts move accordingly, both positions were considered. Finally, because results could differ between populations, we involved two centers: one in the United States and one in the European Union. Materials and Methods Patients Our study was performed in two centers: the Beth Israel Deaconess Medical Center, Boston, Mass (center A) and the Hôpital André Vésale, Montigniesle-Tilleul, Belgium (center B). The relevant ethics committees in each center approved this investigation; however, given local differences in CT equipment, data acquisition was slightly different in the two centers. In center A, data acquisition was exclusively retrospective, and the need for written informed consent Implication for Patient Care Our study results suggest that, if applied to women lying supine and raising their arms above their heads, organ-based tube current modulation would not decrease the radiation dose delivered to external breast tissue by CT; the benefit of scaing with the patient in a prone position is also doubtful. was waived. In center B, data acquisition included a prospective element, as described later; therefore, written informed consent was obtained from all patients. Inclusion criteria in both centers were women referred for routine clinical chest CT who were older than 18 years and had no history of mastectomy or breast implant placement. Repeated chest CT examinations were not included. Center A investigators examined the data in 298 women (mean age, 58 years 6 17 [standard deviation]; range, 18 95 years) who underwent scaing in the supine position between March 2012 and October 2012. Thirty-four women aged 53 years 6 14 (range, 18 84 years) who underwent scaing in the prone position were also included. Forty-five women were excluded because the field of view did not include the entire breast. Center B investigators prospectively included 200 consecutive women (mean age, 64 years 6 14; range, 19 90 years) who underwent a first CT examination of the chest in the supine position between September 2011 and January 2012. Thirty-two women were excluded because they did not meet the inclusion criteria. CT Technique Center A. In picture archiving and communication system (PACS)-archived chest CT examinations, a set of Published online before print 10.1148/radiol.14140694 Content code: Radiology 2015; 274:260 266 Abbreviations: CI = confidence interval OBTCM = organ-based tube current modulation PACS = picture archiving and communication system Author contributions: Guarantor of integrity of entire study, D.T.; study concepts/ study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, S.T., D.E.L., M.S., D.T.; clinical studies, all authors; statistical analysis, S.T., A.A.B.; and manuscript editing, S.T., D.E.L., A.A.B., P.A.G., D.T. Conflicts of interest are listed at the end of this article. Radiology: Volume 274: Number 1 January 2015 n radiology.rsna.org 261

axial reconstructions was displayed in soft-tissue window settings. One chest radiologist (D.E.L.) with 12 years of experience determined the image on which the angles would be measured. This image was chosen to display the maximum dimension of breast parenchyma with at least one of the two nipples, either left or right, visible. This image was then pulled from the PACS (GE Centricity; GE Healthcare) and reloaded onto the scaer (HD 750, 64 row; GE Healthcare). The measurement toolbox of the scaer was used to apply the grid (crosshair) function to the image. This function overlays the image with a crosshair, the center of which represents the isocenter of tube rotation. The image and its overlay were saved and sent back to the PACS for further analysis. Center B. The examinations were performed with two commercially available CT scaers with 16 and 64 detector rows (Somatom Sensation 16 and Definition AS+; Siemens Healthcare, Forchheim, Germany). Women underwent scaing in the supine position with their arms above their heads. To determine the position of the breast tissue with respect to the tube rotation angle, we had to identify the center of tube rotation on the CT images. This center may be distinct from the patient s center because off centering in both the x-axis and y-axis is frequent (7). From the raw data, we therefore reconstructed a series of 5-mm axial images with the largest possible field of view (500 mm) and centered on x = 0 and y = 0 mm, corresponding to the center of tube rotation. This axial series was archived on a server, and measurements were obtained thereafter at a clinical workstation (Syngo- Via; Siemens Healthcare). Thus, no additional studies or radiation exposures were incurred by this portion of the study. In both centers, quality control of the CT devices was performed twice a year. Image Analysis Center A. The center of rotation was determined by using the crosshair placed on the data archived in the PACS. The angles between the y-axis and the tangential line with the ier and outer aspects of the glandular breast tissue were then measured by one radiologist (D.E.L.) using the caliper tool available at the clinical workstation; maximal lateral and anteroposterior chest diameters also were measured. We considered data in 298 women who underwent scaing in the supine position and in 34 women who underwent scaing in the prone position. To verify measurement reproducibility, we chose at random 29 of the women who underwent scaing in the supine position (8) and conducted both inter- and intraobserver repeated measurements, as assessed by two radiologists (D.E.L. and M.S. [with 2 years of experience]). Center B. The center of rotation corresponding to the intersection of diagonal lines was traced on the square CT image. One radiologist with 4 years of experience (S.T.) used a transparency film graduated every degree on the left and right sides of the y-axis that was considered to be 0. He fixed this transparency film to the workstation screen after having verified carefully that its diagonals were superimposed strictly on those appearing on the screen. Because the field of view in successive women was unchanged, the successive corresponding images were displayed the same way as well. With this approach, we determined the angles between the y-axis and the tangential line with the ier and outer aspects of the glandular breast tissue. In addition, maximal lateral and anteroposterior chest diameters were measured. To verify measurement reproducibility, we chose at random 30 women (8) and performed both inter- and intraobserver repeated measurements, as assessed by two observers (S.T. and D.T. [with 28 years of experience]). Determining the Relationship between Measured Angles and OBTCM OBTCM was designed to reduce the tube current by approximately 75% over 80 of the anterior tube rotation (Fig 1). At the edge of this angle, there is a transitional zone of 20 on either side where the tube current is modulated between a maximum of 275% and +25% of the nominal current per section. The amplitude of this modulation depends on technical parameters, including tube rotation time and pitch factor. Therefore, we can assume that the cutoff zone where the tube current passes the nominal current value is situated between 40 and 60 either side of the y-axis of tube rotation, close to 55. However, the precise angle will vary depending on scaing conditions and is not symmetrical (1). The remaining 240 of tube rotation sees an increase of approximately 25% over the nominal tube current. Taking into account the x-ray fan beam, Matsubara et al (6) reported a comparison of skin dose in phantoms; their results showed that the angular position of the transition between decreased and increased skin dose approximated 60 either side of the y-axis, which is in good agreement with the technical description of the manufacturer shown in Figure 1. For our measurements, we therefore used a cutoff greater than 60 left or right of the y-axis for increased tube current and less than 40 left or right of the y-axis for decreased tube current, as proposed by the manufacturer. These cutoffs were used when evaluating the number of breasts that would benefit from dose reduction or dose increase in relation to the iermost and outermost aspects of dense breast tissue. Statistical Analysis Means of quantitative continuous variables between the two groups were compared by using Student t tests. Correlations between continuous variables were quantified with the use of Pearson correlation coefficients. Proportions between the two groups were compared by using the Fisher exact test. Statistical significance was defined as when P,.05. Inter- and intraobserver variability in each center was quantified by calculating the intraclass correlation coefficients for these parameters, together with their 95% confidence intervals (CIs) (9). We used statistical software (SPSS, version 20; IBM, Armonk, NY) in our study. 262 radiology.rsna.org n Radiology: Volume 274: Number 1 January 2015

Figure 1 Figure 1: Graph shows angular tube current modifications as foreseen by the manufacturer for OBTCM CT, with the expected transition between decreased and increased dose represented. Results Supine Position The internal and external angles of both sides and from both centers are displayed in Table 1 and in Figure 2a and 2b. Significant differences were found between centers A and B for all measured angles (P,.001), except for the right internal angle (P =.107). Measured chest diameters are shown in Table 2. Correlations between the angles and chest diameters with patient age at both centers are shown in Table 3. Mean diameters in center B were larger than in center A (+2.3 cm for anteroposterior, +2.5 cm for lateral; P,.001). There was no statistically significant correlation between lateral and anteroposterior chest diameters and the external angles at either center (center A: r ranging from 20.080 to 0.061 and P ranging from.170 to.808; center B: r ranging from 20.001 to 0.066 and P ranging from.699 to.986). The numbers and percentages of women who underwent imaging in the supine position whose external limit of breast Table 1 Measured Angles of Breast Tissue of Both Sides and from Both Centers in Women Lying Supine Angle (degrees) Center A tissue was situated in the increaseddose zone are shown in Table 4. No woman had the external limit of breast tissue within the reduced-dose zone. The internal limit of breast tissue was located within the anterior dose-reduction zone on both sides in 230 (77%) of 298 women in center A and in 133 (67%) of 200 women in center B. Intraobserver variabilities, as assessed with the intraclass correlation Center B Right External 76 6 9 75, 77 83 6 11 81, 84 Internal 29 6 11 28, 30 31 6 15 29, 33 Left External 75 6 9 74, 76 84 6 12 83, 86 Internal 29 6 10 28, 30 35 6 13 34, 37 Note. For right external versus left external, the correlation coefficient was 0.727 (P,.001) for center A and 0.485 (P,.001) for center B; for right internal versus left internal, the correlation coefficient was 0.541 for center A (P,.001) and 0.375 (P,.001) for center B. coefficient, were 0.995 (: 0.993, 0.997) and 0.992 (: 0.989, 0.995) in centers A and B, respectively. Interobserver variabilities were 0.977 (: 0.897, 0.991) and 0.993 (95% CI: 0.989, 0.995) in centers A and B, respectively. Prone Position External and internal angles were 66 6 11 (: 62, 69 ) and 22 6 Radiology: Volume 274: Number 1 January 2015 n radiology.rsna.org 263

Figure 2 Figure 2: (a c) Graphic representations of a woman in relation to the center of the CT scaer are based on average chest diameter compared with a field of view of 500 mm. The mean and the standard deviation of the ier and outer limits of breast positions are represented in the ier circle. Color-coded areas represent the anterior dose-reduction and the posterior dose-increase zones, as well as the intermediate zones. The assumed average 55 cutoff zone between decreased versus increased zones is also shown. Images were obtained from (a) supine data from center A, (b) supine data from center B, and (c) prone data from center A. A = anterior, P = posterior, R = right, L = left. proportion of breast tissue between increased versus reduced-dose zones was 0.34 and 0.35 for the right and left breasts, respectively. The internal limit of breast tissue was located within the anterior dose-reduction zone on both sides in 33 (97%) of 34 women. 9 (: 19, 25 ) for the right side and 67 6 9 (: 64, 70 ) and 20 6 12 (: 16, 24 ) for the left side. These angles are demonstrated in Figure 2c. Twenty-eight (82%) of 34 women had at least one breast s external limit lying in the increased-dose zone. No woman had the external limit of breast tissue within the reduced-dose zone. The ratio of angular Discussion Our chest CT study results show that no woman had the entirety of breast tissue within the reduced-dose zone in the supine or prone position because nearly all women (.99%) had at least a part of their breast tissue within the increased-dose zone. Although significant differences in angle measurements and in diameters existed between the two centers located on two 264 radiology.rsna.org n Radiology: Volume 274: Number 1 January 2015

Table 2 Measured Chest Diameters Diameter (cm) Center A Center B P Value Anteroposterior 21.4 6 2.2 21.1, 21.6 23.7 6 2.8 23.3, 24.1,.001 Lateral 31.4 6 3 31.1, 31.8 33.9 6 4.6 33.3, 34.6,.001 Table 3 Correlations between Angles and Diameters and Patient Age in Women Lying Supine Center and Statistic Right Angle Left Angle Diameter External Internal External Internal Anteroposterior Lateral Center A Correlation coefficient 20.065 0.198 20.023 0.288 0.312 20.143 P value.267.001.690,.001,.001.013 Center B Correlation coefficient 0.269 0.260 0.271 0.334 0.281 20.044 P value,.001,.001,.001.001,.001.535 Table 4 Number of Women in the Supine Position with the External Limit of Breast Tissue in the Increased-Dose Zone Center and P Value Right Breast Left Breast At Least One Breast Center* A (n = 298) 290 (97.3) 281 (94.3) 295 (99.0) B (n = 200) 199 (99.5) 199 (99.5) 200 (100) P value for center A vs center B.092.002.278 * Data are numbers of patients, with percentages in parentheses. different continents, the proportions of women with breast tissue exposed to an increased dose was similar between centers. Therefore, our study results suggest that in women lying supine and raising their arms above their heads, OBTCM does not decrease the radiation dose delivered to external breast tissue. The benefit of scaing with the patient in the prone position is also doubtful. Our study results show that with the patient in the prone position, although the breasts appear to be more centrally located, there is at least a part of the breast tissue that lies in the increased-dose zone. We were able to detect a relationship between angular breast position and age in women in the supine position but not with chest diameters. This finding indicates that the breast tissue tends to move laterally with age but that body shape has no influence, which increases the relative risk of older women having larger portions of breast tissue in the zone of increased radiation dose. Although there is a positive correlation between age and angular breast position, age should not be used for selection because no woman had her breasts entirely within the reduced-dose zone in our study. Thus, our results do not allow the determination of selection criteria for the use of OBTCM. Even if the prone position brings the breasts to a more medial position, which could be beneficial for the use of OBTCM, our results raise doubts about this benefit because the external part of the breast, which is still in the increased-dose zone, is where breast cancer most frequently develops (10). Other researchers have investigated the effect of OBTCM on organ dose and on image quality in phantoms (1,4 6). The position of breast tissue in an anthropomorphic phantom is located anteriorly and medially as in an erect adult woman. Dose reductions to anteriorly located organs have been measured, but the position of the phantom breast tissues does not represent the position observed in our study with women lying supine. Lungren et al (5) addressed chest organ doses in anthropomorphic phantoms and breast position in 100 consecutive women, and they concluded that a reduced dose potentially can be delivered to the breasts. Our methods and results differ from those of that investigation because those authors considered the anterior 120 of rotation as the dose-reduction angle and did not consider the transition angular zone of 20 on each side. In addition, they did not discriminate between internal and external angular positions, as well as the woman s center and that of tube rotation, exposing them to off-centering bias. Finally, they did not consider as important that the external part of the breasts, where most glandular tissue lies, was outside of the reduced-dose zone. Our study was composed of two relatively large samples from two continents. Even though the measured angles differed between them, the results from both centers were in agreement that external breast limits are located in the higher-dose zone. Our study had several limitations. First, we considered dense breast tissue within the fatty breast tissue to be the limit of glandular tissue. We therefore could not discriminate formally between more fibrous breast tissue and breast tissue with a greater glandular component. This difference warrants further study to determine the distribution within breast tissue of the composition and distribution of radiation-sensitive tissues and how this can affect such Radiology: Volume 274: Number 1 January 2015 n radiology.rsna.org 265

dose-reduction techniques. Second, we did not measure the dose delivered to the breasts and other organs; we measured only the angular breast positions. Thus, we could not demonstrate definitely that OBTCM increases the dose to the breast and possibly the overall effective dose to the chest, but we can speculate on these higher doses on the basis of the manufacturer s technical instructions and previous articles in which the authors addressed dosimetry (1,5,6). Third, we did not evaluate any parameters reflecting image quality because OBTCM was not switched on in the patients in our study. Fourth, the series of women in our study who were lying prone was much smaller than the two other series of women who were lying supine. However, this difference reflects current practice, because most chest CT examinations are performed in patients lying supine. Fifth, we evaluated the angular position of breasts on only one section the section in which the highest angular position was measured. This issue warrants further study, including a volumetric approach of breast position. Sixth, these results may apply only to the particular types of CT scaers used in our study; at this time, we do not know how other scaers may function in this mode. Finally, we did not collect data from a population with known smaller chest diameters, including children and Asians. In conclusion, our data show that the breast angular position of almost all women is higher than the angular limit of reduced versus increased radiation dose during OBTCM and that no woman, regardless of supine or prone position, has all breast tissue within the reduced-dose zone, raising substantial concerns regarding the potential of OBTCM to reduce CT irradiation of breast tissue successfully. Disclosures of Conflicts of Interest: S.T. disclosed no relevant relationships. D.E.L. disclosed no relevant relationships. M.S. disclosed no relevant relationships. A.A.B. Activities related to the present article: none to disclose. Activities not related to the present article: is a consultant for Spiration and Olympus; is on the speakers bureau of ATS; receives royalties from Elsevier. Other relationships: none to disclose. P.A.B. disclosed no relevant relationships. D.T. disclosed no relevant relationships. References 1. Duan X, Wang J, Christner JA, Leng S, Grant KL, McCollough CH. Dose reduction to anterior surfaces with organ-based tube-current modulation: evaluation of performance in a phantom study. AJR Am J Roentgenol 2011;197(3):689 695. 2. Wang J, Duan X, Christner JA, Leng S, Grant KL, McCollough CH. Bismuth shielding, organ-based tube current modulation, and global reduction of tube current for dose reduction to the eye at head CT. Radiology 2012;262(1):191 198. 3. Hoang JK, Yoshizumi TT, Choudhury KR, et al. Organ-based dose current modulation and thyroid shields: techniques of radiation dose reduction for neck CT. AJR Am J Roentgenol 2012;198(5): 1132 1138. 4. Wang J, Duan X, Christner JA, Leng S, Yu L, McCollough CH. Radiation dose reduction to the breast in thoracic CT: comparison of bismuth shielding, organ-based tube current modulation, and use of a globally decreased tube current. Med Phys 2011;38(11):6084 6092. 5. Lungren MP, Yoshizumi TT, Brady SM, et al. Radiation dose estimations to the thorax using organ-based dose modulation. AJR Am J Roentgenol 2012;199(1): W65 W73. 6. Matsubara K, Sugai M, Toyoda A, et al. Assessment of an organ-based tube current modulation in thoracic computed tomography. J Appl Clin Med Phys 2012;13(2): 3731. 7. Gudjonsdottir J, Svensson JR, Campling S, Brean PC, Jonsdottir B. Efficient use of automatic exposure control systems in computed tomography requires correct patient positioning. Acta Radiol 2009;50(9): 1035 1041. 8. Fisher RA, Yates F. Statistical tables for biological, agricultural and medical research. London, England: Oliver & Boyd, 1963. 9. Bankier AA, Levine D, Halpern EF, Kressel HY. Consensus interpretation in imaging research: is there a better way? Radiology 2010;257(1):14 17. 10. Lee AHS. Why is carcinoma of the breast more frequent in the upper outer quadrant? a case series based on needle core biopsy diagnoses. Breast 2005;14(2):151 152. 266 radiology.rsna.org n Radiology: Volume 274: Number 1 January 2015