| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Institute of Radiology, University of Udine, Via Colugna 50, Udine 33100,
Italy.
2 Department of Radiology, Vita-Salute University, San Raffaele Hospital, Milan,
Italy.
3 Department of Medical Statistics, Faculty of Medicine, University of Udine,
Udine, Italy.
4 Department of Diagnostic Imaging, Istituto Policlinico San Donato, Milan,
Italy.
5 Institute of Radiology, University of Ancona, Ancona, Italy.
6 Department of Diagnostic Imaging and Interventional Radiology, University of
Tor Vergata, Rome, Italy.
7 Department of Senology, Policlinico Universitario Bari, Bari, Italy.
Received January 24, 2005;
accepted after revision July 15, 2005.
Address correspondence to M. Bazzocchi.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. One hundred twelve patients with
mammographically detected microcalcifications with BI-RADS category 5
(n = 78) or 4 (n = 34) lesions were studied at 17 centers a
using 3D gradient-echo dynamic coronal technique (
3 mm thickness) and 0.1
mmol/kg of gadoteridol. A pathologic sample was obtained in all cases.
Agreement between the major diameter measured on mammography, MRI, or both and
the major diameter measured at pathologic examination was calculated in 62
cases.
RESULTS. Of the 112 lesions, pathologic examination revealed 37 benign lesions, 33 ductal carcinoma in situ (DCIS), and 42 invasive carcinomas. The specificity of MRI for benign lesions was 68%. Considering the subgroups of calcifications alone and calcifications associated with masses, the specificity values became 79% and 33%, respectively. The sensitivity of MRI for DCIS was 79%. Analysis of the two subgroups showed sensitivity values of 68% for calcifications alone and of 1% for calcifications associated with masses. The sensitivity for invasive carcinomas was 93%. Analysis of the two subgroups showed sensitivity values to be 92% for calcifications alone and 94% for calcifications associated with masses. Considering the overall results, the sensitivity of MRI was 87%; specificity, 68%; positive predictive value, 84%; negative predictive value, 71%; and accuracy, 80%. Considering the subgroups of calcifications alone and calcifications associated with masses, the sensitivity values became 80% and 97%; the positive predictive values, 86% and 82%; the negative predictive values, 71% and 75% (95% confidence interval [CI], 0.19-0.99); and the accuracy values, 80% and 82% (95% CI, 0.66-0.92), respectively. An odds ratio (OR) of 13.54 (95% CI, 5.20-35.28) showed a raised risk of malignant breast tumor in subjects with positive MR examination of mammographically detected suspicious clusters of microcalcifications. The statistical analysis on each subgroup showed an OR of 15.07 (95% CI, 4.73-48.08) for calcifications alone and an OR of 14.00 (95% CI, 1.23-158.84) for calcifications associated with masses. Any significant improvement in the predictive ability of dynamic MRI depending on the extent of calcifications on mammography was not proved. Considering the 62 cases of proved malignancy with measured maximal diameter at pathologic examination, both mammography and MR examination seem to overestimate tumor extent.
CONCLUSION. The not-perfect sensitivity of MRI (87%), when applying our interpretation criteria and imaging sequences, is a crucial point that prevents us from clinical use of MRI in the diagnosis of mammographically detected microcalcifications.
Keywords: breast • breast carcinoma • breast microcalcifications • mammography • MRI
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We report the results of a multicenter trial aimed at testing dynamic contrast-enhanced breast MRI in evaluating patients with mammographically detected suspicious microcalcifications.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
15 mT/m gradients, and bilateral synchronous breast coil. The
ethics committee of each center approved the research. Patients gave informed
written consent to the protocol and to the use of the data obtained from
statistical analysis and their publication.
Patients
Patients 18 years old or older of any race with suspicious
microcalcifications classified as BI-RADS category 4 or 5
[16] seen on mammograms at
each enrolling center, associated or not with an opacity, were eligible for
recruitment. Exclusion criteria were absolute contraindications to MR,
pregnancy or breastfeeding, severe renal failure, known hypersensitivity to
gadolinium chelates, inclusion in other clinical trials during the month
before enrollment, and clinical status limiting data reliability.
Mammography
Mammography was performed with standard high-quality methods: a two- or
three-view examination with targeted compression and magnification for fine
analysis of microcalcifications.
MRI
Dynamic MRI was performed on the 7-14th day of the menstrual cycle in
fertile women, within 30 days after mammography. A long-line venous access was
obtained before starting the examination. An unenhanced 3D gradient-echo
coronal sequence was used with a 3 mm thickness partition without a gap.
The minimum TR allowed by the MR equipment, and the TE was chosen to avoid the
phase opposition between fat and water.
An IV bolus injection of gadoteridol (ProHance, Bracco Imaging) at the
standard single dose of 0.1 mmol/kg of body weight was administered at the
rate of 2 mL/sec, using an automatic injector. The contrast-enhanced sequence
was acquired starting 12 sec after the beginning of contrast injection and was
repeated with a temporal resolution of 90 sec, up to 8 min after the
beginning of contrast injection with at least five acquisitions.
Postprocessing included temporal subtraction (contrast-enhanced minus
unenhanced), maximum intensity projections, multiplanar reconstructions, and
dynamic signal intensity-time or percent enhancement-time curves (or both)
based on regions of interest (3 x 3 pixels) positioned within the
suspected lesion on the subjectively recognized areas of maximal
enhancement.
Surgical Biopsy
After a hookwire was positioned under mammographic guidance, pathologic
diagnosis was obtained in all patients after surgical biopsy, within 2 weeks
after MRI. All surgical specimens underwent radiographic examination to
confirm the excision of microcalcifications. The lesions were then
pathologically classified as benign or malignant, the latter of which was
subdivided in DCIS, invasive ductal carcinoma (IDC), and invasive lobular
carcinoma (ILC). Lobular carcinoma in situ (LCIS), atypical lobular
hyperplasia, and atypical ductal hyperplasia were considered benign lesions,
even though the presence of these lesions increases the risk of developing a
malignant lesion in both breasts
[17].
Analysis of Cases
From 1998 to 2000, 17 centers enrolled 174 patients. Two levels of check
were activated to define the cases with comparable mammography, MRI, and
pathologic data: cases with planned but not performed open surgery; with
incomplete mammography, MRI, or pathology; and cases with important imaging
artifact (e.g., patient's movement between MRI unenhanced and
contrast-enhanced scans) were excluded from analysis. At the central unit
(off-site), two experienced radiologists checked all the examinations with
respect to the protocol and the quality of the images. As a consequence, 62
patients were excluded from the analysis, and 112 cases remained for complete
mammographic, MRI, and pathologic correlation.
For these 112 cases, an off-site interpretation of the mammograms and MR images was performed by consensus of two experienced radiologists, aware of the entry criterion (presence of suspicious micro-calcifications on the mammograms) and blinded to all of the patients' data and to results of pathologic examination. The two reviewers by consensus interpreted the MR examinations on the printed films supplied by each center, under guidance of the mammographic image; they integrated morphologic and dynamic data into the 0- to 8-point score proposed by Fischer et al. [18] (Table 1). Comparison of breast findings at pathologic examination and MRI was performed off-site by a third radiologist who matched MR and pathologic reports, creating a data-base for the data analysis, considering only the area of interest for microcalcification.
|
The maximal diameter of malignant lesions was assessed at both mammography and MRI and compared with that obtained at pathologic examination, which was used as a standard of reference, when available.
Statistical Analysis
Because of the likely pathologic differences between calcifications alone
and calcifications associated with masses, the statistical analysis, when
necessary, was performed on each subgroup.
To describe dynamic MRI in evaluating mammographically detected suspicious microcalcifications, we used the sensitivity, specificity, and positive and negative predictive values with their respective 95% CIs.
To test the ability of dynamic MRI to be used to predict pathologic diagnosis, we calculated the odds ratio (OR) for pathologically detected breast tumors associated with positive MR examination evaluated by means of the Fischer's score.
To assess the better or worse predictive ability of dynamic MRI depending on the extent of calcifications on mammography, we used a logistic regression model.
To define whether the agreement in terms of lesion diameter measures between mammography or MR and pathologic examination was clinically acceptable, we used the Bland-Altman analysis for all cases of pathologically proven malignancy true-positive on one or both of the imaging techniques and with the extent measured and available in the pathologic report.
The reliability of the lesion diameter measurements obtained with mammography and those obtained with MRI was assessed by the interclass correlation coefficient [19, 20].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Benign Lesions
The pathologic diagnosis of benign lesion was made in 37 patients:
fibrocystic disease in 16 patients, typical ductal hyperplasia in four,
fibroadenomas in five, adenosis in one, atypical lobular hyperplasia in two,
atypical ductal hyperplasia in two, papillomatosis in one, LCIS in one, and
mixed benign lesions in five. The size of the lesions ranged between 3 and 100
mm (mean, 24 mm) on mammography and between 5 and 100 mm (mean, 27.57 mm) on
MRI.
MRI results were as follows: 23 (62.1%) of 37 lesions did not show any enhancement after contrast agent injection, whereas 14 lesions (37.8%) showed enhancement. Of those 14 enhancing lesions, two were classified as benign (both lesions, score of 2 according to the system described by Fischer et al. [18]) and 12 as suspicious (three lesions, score of 3; four lesions, score of 4; three lesions, score of 6; one lesion, score of 7; and one lesion, score of 8). The 12 false-positive cases on MRI were four cases of fibrocystic diseases (Figs. 1A, 1B, 1C, 1D and 1E), two cases of papillomatosis, and one case for each of the following: typical ductal hyperplasia, sclerosing adenosis, ductal sclerosis, apocrine papillary hyperplasia, fibroadenoma, and LCIS. Six of the 12 false-positive cases presented with clusters of microcalcifications and six with a cluster of microcalcifications associated with an opacity.
|
|
|
|
|
DCIS
The pathologic diagnosis of DCIS was made for 33 lesions: comedo-type DCIS
in 18 (54.5%), noncomedo-type DCIS in 11 (33.3%), and nonspecified DCIS in
four (12.1%). The size of the lesions ranged between 3 and 150 mm (mean, 31.42
mm) on mammography and between 4 and 110 mm (mean, 34.68 mm) on MRI.
On MRI, four (12.1%) of 33 lesions did not show any enhancement after contrast agent injection (Figs. 2A, 2B and 2C), whereas 29 lesions (87.8%) showed enhancement (Figs. 3A, 3B and 3C). Of those 29 enhancing lesions, three were classed as benign (one lesion, score of 1; two lesions, score of 2) and 26 as suspicious (four lesions, score of 3; one lesion, score of 5; eight lesions, score of 6; 12 lesions, score of 7; and one lesion, score of 8).
|
|
|
|
|
The sensitivity of the Fischer score for the overall DCIS was 0.79 (95% CI, 0.61-0.91). Analysis of the two subgroups showed sensitivity values of 0.68 (0.45-0.86) for calcifications alone and 1.00 (0.72-1) for calcifications associated with masses.
Invasive Carcinomas
The pathologic diagnosis of invasive carcinoma was made for 42 lesions: IDC
in 36 lesions (85.7%) and microinvasive carcinoma in six (14.3%). The size of
the lesions ranged between 5 and 130 mm (mean, 39.40 mm) on mammography and
between 6 and 100 mm (mean, 44.4 mm) on MRI.
The MRI results were two (4.8%) of 42 lesions did not show any enhancement, whereas 40 (95.2%) enhanced (Figs. 4A, 4B, 4C and 4D). Of those 40 enhancing lesions, one was classified as benign (score of 2) and 39 were classified as suspicious (six lesions, score of 4; four lesions, score of 5; 13 lesions, score of 6; and 16 lesions, score of 7). The two MR false-negative cases with no enhancement were one mixed ILC-LCIS and one mixed DCIS-IDC of 5 and 20 mm, respectively, and both presented mammographically as a cluster of microcalcifications only; the invasive component was a minimal part of those lesions. The MR false-negative case with a Fischer score of 2 was a mixed DCIS-invasive tubular carcinoma of 20 mm that presented on mammography as a cluster of microcalcifications associated with an opacity.
|
|
|
|
Overall Mammography and MR Diagnostic Performance and Tumor Extent Evaluation
Considering the overall DCIS and invasive carcinoma results, the
sensitivity and the specificity of mammography were, of course, equal to 1 and
zero, respectively, and the positive predictive value was 0.67 (95% CI,
0.57-0.76), which, in our study group, agrees with the accuracy value.
The sensitivity of the Fischer score was 0.87 (95% CI, 0.77-0.93), the specificity was 0.68 (0.50-0.82), the positive predictive value was 0.84 (0.74-0.92), the negative predictive value was 0.71 (0.54-0.85), and the accuracy was 0.80 (0.72-0.87).
With an OR of 13.54 (95% CI, 5.20-35.28), we can infer that there is a raised risk of malignant breast cancer in subjects with positive findings according to the Fischer score, which is obtained by evaluating images.
Considering the subgroups of calcifications alone and calcifications associated with masses, the sensitivity values became 0.80 (95% CI, 0.66-0.91) and 0.97 (0.82-1); the positive predictive values, 0.86 (0.72-0.95) and 0.82 (0.65-0.93); the negative predictive values, 0.71 (0.52-0.86) and 0.75 (0.19-0.99); and the accuracy values 0.80 (0.69-0.88) and 0.82 (0.66-0.92), respectively.
The statistical analysis on each subgroup showed that that risk of malignant breast cancer in subjects with positive results according to the Fischer score, which is obtained by evaluating images, became an OR of 15.07 (95% CI, 4.73-48.08) for calcifications alone and an OR of 14.00 (1.23-158.84) for calcifications associated with masses.
The logistic regression analyses performed on the overall 112 lesions, the 74 calcifications alone, and the 38 calcifications associated with masses did not show any significant improvement in the predictive ability of dynamic MRI depending on the extent of calcifications on mammography.
Moreover, considering the cases of pathologically proven malignancies for which the extent was available in the pathologic report (n = 62), the mean diameter was 29.80 ± 24.50 (SD) mm at the pathologic examination, 33.71 ± 26.64 mm on mammography, and 37.13 ± 27.19 mm on MRI.
The mean difference in lesion size between mammography and pathologic examination was 3.91 mm, with a 95% CI of -1.39 to 9.21. Thus, mammography tends to give a higher reading for lesion size, but the size reading based on mammography does not significantly differ from that based on pathology.
The mean difference between MRI and pathologic examination was 7.33 mm with a 95% CI of 1.47-13.18. Thus, MRI tends to give a size reading that is significantly greater than the pathologic one.
Regarding the Bland-Altman analysis, because the differences between methods were proportional to their mean, a logarithmic transformation improved the picture, and we applied the analysis to the transformed data.
Figure 5 shows a plot of the difference between the logarithms of the major lesion's diameter as measured on mammography and at pathologic examination against their mean.
|
Even though the measurements were not significantly different, the limits of agreement are not small enough for us to be confident that mammography can be used for measuring lesion diameter.
Figure 6 shows a plot of the difference between the logarithms of the major lesion's diameter as measured on MRI and at pathologic examination against their mean.
|
|
The limits of agreement are definitely too wide to use MRI for measuring lesion diameter. Moreover, the interclass correlation coefficients between pathologic examination and mammography (0.80, p < 0.001) and between pathologic examination and MRI (0.75, p < 0.001) showed a higher reliability for mammography.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several authors have emphasized the value of breast MRI not only in the diagnosis of invasive carcinoma, but also in the differentiation of in situ ductal carcinomas with neoangiogenesis in the stroma surrounding tumoral ducts as a marker of potential dissemination [21]. Furthermore, Gilles et al. [22] emphasized MR evaluation of the real DCIS extent with interesting levels of accuracy (75-80%), opening a possible role for MRI in surgical planning for management of breast malignancies associated with microcalcifications [22].
Breast MRI is based on the increase of signal intensity in malignant lesions after IV contrast agent administration due to the presence of more and larger vessels with higher permeability and to an increase of interstitial extravascular space [23]. In fact, tumor cells have been shown to release angiogenesis factors that result in the diffusion and growth of preexisting capillaries and the creation of new vessels with discontinuous endothelium. As a consequence, a faster transit of contrast agent in the extravascular space inside the malignant lesion is observed [24]. However, neoangiogenesis is not the same for all malignant breast lesions: It is more pronounced in IDC and is not always present in DCIS and lobular carcinomas, in which cells present a growing model "in single file" and the preexisting capillaries are not able to support all the tumoral cells [25].
The increase of signal intensity in lesions is a multifactorial process. Vessel density, vascular permeability, and interstitial volume and biochemical structure influence perfusion and diffusion of contrast agent inside and around the tumor. Aggressive lesions usually have high vessel density resulting in early and strong enhancement. On the other hand, several malignant lesions have a low density of vessels—in particular, preinvasive or in situ cancers. Furthermore, some degree of neoangiogenesis seems to be a fundamental prerequisite for tissue invasion, whereas DCIS presents a variable degree of vascularization [25].
Therefore, neoangiogenesis would explain the reported high sensitivity of contrast-enhanced MRI in detecting invasive breast cancers (94-100%) [5] and the relatively reduced sensitivity in detecting DCIS (45-100%) [6-15, 26]. The wide range of sensitivity reported for DCIS probably reflects the relatively small numbers of cases, ranging from 13 to 36 patients in the different series, and the variable imaging techniques. For example, better results were achieved using a single dedicated breast surface coil [7, 11, 13, 22] compared with the bilateral breast coil [12], even though routinely the use of a bilateral coil is preferred because it allows simultaneous imaging of both breasts, which is relevant in particular for staging purposes. Moreover, higher sensitivity has been obtained using RODEO (rotation delivery of excitation off-resonance) sequences and 3D sequences [9, 26] rather than using 2D sequences. Now, high-spatial-resolution 3D techniques should be used to provide large area coverage with thin (< 3 mm) sections and no intersection gap.
The purpose of this multicenter trial was to evaluate the role of breast MRI in the detection and characterization of clusters of microcalcifications seen on mammograms. Few articles have been published on this topic, because some authors have been reluctant to include these kinds of lesions in research studies on the basis of their belief that these findings are not an indication for breast MRI [5]. The few published articles present discordant data, probably because of the difference in techniques and analysis criteria [10-13, 15].
Our results showed a relatively low sensitivity (0.87) of breast MRI for the detection of breast cancers associated with microcalcifications. Moreover, the study confirms the relatively low positive predictive value of suspicious microcalcifications on mammograms (0.67) in BI-RADS category 4 or 5 lesions. Due to the entry criterion (patients with a suspicious cluster of mammographically detected microcalcifications), the combination of high sensitivity and low specificity of mammography was expected. MRI showed a relatively good specificity (0.68), which is well represented by the 23 (79%) of 29 benign lesions without enhancement on MRI that were scored as BI-RADS category 4 or 5 lesions on mammograms. This is due to the absence of or very low neoangiogenesis in most benign breast lesions [23-25].
However, the not-perfect sensitivity of MRI—87%—is a crucial point that prevents us from clinical use of MRI in the diagnosis of mammographically detected microcalcifications. In fact, MRI was negative in seven (21.2%) of 33 cases of DCIS (in four cases, no enhancement at all; and in three cases, benign enhancement based on the Fischer's score) and in three (7.14%) of 42 invasive carcinomas (in two cases, no enhancement at all; and in one case, benign enhancement based on the Fischer's score). This difference in sensitivity was expected on the basis of less pronounced neoangiogenesis in intraductal tumors [24, 25]. The negative MRI findings in three cases of invasive cancers have a similar explanation: One was a mixed LCIS-ILC and two were mixed DCIS-invasive carcinoma.
Analysis of the two subgroups of lesions, presenting as a cluster of microcalcifications alone or as clusters of microcalcifications associated with opacities, shows interesting results. The sensitivity of breast MRI increases significantly for the second subgroup especially for DCIS (1.00 vs 0.68, respectively). This result suggests that neoangiogenesis is more likely present in lesions with an opacity associated with microcalcifications. All seven false-negative DCIS and two of the three false-negative invasive carcinomas presented on mammograms as a cluster of microcalcifications without an opacity. On the other side, our results showed that the sensitivity of breast MRI is not correlated with the extent of microcalcifications on mammography; probably, a small cluster of microcalcifications from an invasive carcinoma is more likely seen on breast MRI than a large cluster of microcalcifications from a DCIS with low angiogenesis.
Our trial confirms the relatively low specificity and overall accuracy obtained by other authors who have studied a sufficiently large number of clustered microcalcifications using contrast-enhanced MRI. Gilles et al. [11] studied 172 cases and obtained a 51% specificity with a 70% overall accuracy, and Westerhof et al. [12] studied 63 cases and obtained a 72% specificity with a 56% overall accuracy. Thus, our 68% specificity with an 80% overall accuracy is a good result, probably due to technologic improvements—a 3D sequence was not used in the two past studies—and to combined morphologic-dynamic evaluation. However, the overall sensitivity in our study is not more than 87% (combining 93% for invasive and 79% for DCIS), which is different from 95% reported by Gilles et al. (combining 97% and 90%) and 45% reported by Westerhof et al. (combining 100% and 36%); these differences are probably due to the use of different diagnostic criteria. Again, the overall sensitivity rises to 97% when considering lesions presenting as microcalcifications associated with an opacity compared with 80% for lesions presenting as a cluster of microcalcifications alone.
According to our experience with the use of up-to-date 3D sequences and combined morphologic-dynamic evaluation, negative breast MRI findings should not be considered a sure marker of benignancy and should indicate that a reliable workup, particularly for lesions presenting mammographically as clusters of microcalcifications alone, is needed. This trial emphasizes the limitations of breast MRI in characterizing suspicious microcalcifications detected on mammograms. Breast MRI cannot replace percutaneous or surgical biopsy, which gives pathologic information.
Another consideration comes from the real utility of quantitative analysis, which is part of the Fischer's score that we used, in these lesions presenting as a cluster of microcalcifications with a high prevalence of DCIS. In fact, the accurate placement of a region of interest over the area of most rapid and intense enhancement is critical, and interobserver variability and bias in placement have been reported. This is more critical for DCIS, which often presents with linear and branched enhancement, given that the correct placement of the region of interest may not be as accurate as that for solid masses. The presence of nonenhancing cancers in a majority of DCIS and in some invasive carcinomas remains a problem for breast MRI that is independent from any quantitative analysis and probably is mainly related to histologic type (low neoangiogenesis).
In our experience evaluating the extent of malignancies associated with microcalcifications, MRI was not more accurate than mammography. Both imaging techniques seem to overestimate the lesion diameter if the maximal diameter at pathologic examination is used as the standard of reference. However, this point is a crucial one that deserves clarification in future studies. In fact, a maximal in-plane diameter is not a good method for evaluating the tumor extent, which is a 3D volume parameter. Moreover, when ex vivo measurements are compared with in vivo measurements, a reduction in volume and diameter must be expected. In terms of practical use, for tumors detected because of microcalcifications on mammography, MRI is not able to present a map of the real extent of the lesion more reliably than mammography. This result is in agreement with the results obtained by Westerhof et al. [12] who did not note any change of surgical management based on MRI findings in 63 patients with mammographically detected suspicious microcalcifications. A similar experience was reported by Boetes et al. [27]; those researchers evaluated the DCIS component associated with IDC and MRI underestimated tumor size.
All our results can be explained by the relatively high fraction of in situ carcinomas in a population of patients enrolled because of the presence of suspicious microcalcifications: A lower neoangiogenesis reduces the ability of MRI to be used successfully to detect and characterize lesions and evaluate local extent.
We are aware of the debate about the prognostic value of MRI [28] with possible future developments in differentiating the degrees of aggressiveness on the basis of MR patterns, but to date it is difficult to apply this point of view for evaluation of the clinical use of MRI in mammographically detected suspicious microcalcifications.
In conclusion, for women presenting with suspicious calcifications on mammography, particularly when that is the only mammographic finding, breast MRI does not appear to be useful in the presurgical evaluation when applying our interpretation criteria and imaging sequences.
Acknowledgments
The Italian Trial for Breast MRI of Microcalcifications was promoted by the
Italian Association for Medical Radiology (Sections of Senology and Magnetic
Resonance. Enrolling centers were as follows: Udine, University Policlinico
(M. Bazzocchi); Ferrara, Ospedale S. Anna (S. Corcione); L'Aquila, University
Ospedale S. Maria (G. Masciocchi); Rome, Cattolica University, Policlinico
Gemelli (G. Pastore); Trieste, University Ospedale Gattinara (R. Pozzi
Mucelli); Aviano, Centro di Riferimento Oncologico (S. Morassut); Milan,
Istituto Nazionale Tumori (R. Musumeci); Siena, University Policlinico Le
Scotte (P. Stefani); Ancona, University Ospedale Torrette (G. M. Giuseppetti);
Rome, Ospedale Fatebenefratelli (A. Orlacchio); S. Giovanni Rotondo,
Scientific Institute (M. Cammisa) and Chieti University (L. Bonomo); Milan,
Vita-Salute University, Ospedale S. Raffaele (P. Panizza); Firenze, University
Ospedale Careggi (N. Villari) and Center for Study and Prevention of Cancer
(D. Morrone); Rome, Tor Vergata University (G. Simonetti); Milan, European
Institute of Oncology (M. Bellomi); Milan, Ospedale S. Carlo (D. Vergnaghi);
and Reggio Emilia, Ospedale (A. Troiso).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. Lee, M. Pintilie, D. C. Hodgson, P. E. Goss, and M. Crump Screening mammography for young women treated with supradiaphragmatic radiation for Hodgkin's lymphoma Ann. Onc., January 1, 2008; 19(1): 62 - 67. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
