Common Occurrence of Benign Liver Lesions in Patients With Newly Diagnosed Breast Cancer Investigated by MRI for Suspected Liver Metastases

Transcription

1 JOURNAL OF MAGNETIC RESONANCE IMAGING 10: (1999) Original Research Common Occurrence of Benign Liver Lesions in Patients With Newly Diagnosed Breast Cancer Investigated by MRI for Suspected Liver Metastases Tara C. Noone, MD, 1 Richard C. Semelka, MD, 1 * N. Cem Balci, MD, 1 and Mark L. Graham, MD 2 The purpose of this study was to determine the prevalence of benign liver lesions in patients with breast cancer who are referred to magnetic resonance (MR) imaging for suspected breast cancer metastases at initial presentation. The original MR imaging reports of consecutive patients with breast cancer were reviewed; these patients had undergone MR imaging at our institution to investigate for suspected breast cancer liver metastases, at initial presentation between April 1993 and May Determination of the presence of benign and malignant liver lesions in each patient was made, as well as their relative frequencies. Diagnostic accuracy of MR imaging was evaluated by correlation with histologic specimens (5 patients) and imaging follow-up (27 patients). Thirty-four patients with newly diagnosed breast carcinoma were evaluated with MR imaging. A total of 11 (32%) of these patients had benign lesions only. Of 21 (62%) total patients who had malignant liver lesions, 19 had breast cancer metastases (2 had coexistent benign lesions), 1 had metastatic carcinoid, and 1 had hepatocellular carcinoma. No liver lesions were detected in two patients (6%). In one patient with biopsy-proven subcentimeter breast metastases, no focal lesions were shown on MR imaging. No other diagnostic errors in classification of liver lesions by MR imaging occurred, as shown by clinical correlation and imaging follow-up in all patients. True positive detection of malignant liver lesion was 20/ 21, true negative was 13/13, false positive was 0/13, and false negative was 1/21, for a sensitivity of 95% and a specificity of 100% for the detection of malignant liver lesions. Benign liver lesions are common in breast cancer patients suspected clinically of having liver metastases. Benign lesions alone were observed in one-third of our patients. The high diagnostic accuracy of MR imaging in the evaluation of hepatic lesions underscores the value of this technique for baseline investigation of breast cancer patients with clinically suspected liver metastases, particularly patients in whom treatment approaches are dramatically affected by the presence of liver metastases. J. Magn. Reson. Imaging 1999;10: Wiley-Liss, Inc. 1 Department of Radiology, the University of North Carolina at Chapel Hill, Chapel Hill, North Carolina Department of Medicine, the University of North Carolina at Chapel Hill, Chapel Hill, North Carolina *Address reprint requests to: R.C.S., Department of Radiology, UNC Hospitals, 101 Manning Drive, Chapel Hill, NC Received January 15, 1999; Accepted April 29, Wiley-Liss, Inc. 165 BREAST CARCINOMA IS RESPONSIBLE for the greatest number of cancer deaths in American women, annually (1,2). Recent improvements in screening and detection have resulted in earlier local and systemic therapy. Increased utilization of adjuvant therapies also has contributed to a decline in breast cancer mortality in recent years (2). Allocation of patients to a particular treatment arm of adjuvant therapy protocols often depends on the detection of stage IV disease (3,4). Therefore, accurate detection or exclusion of liver metastases following initial diagnosis may determine eligibility for potentially curative therapy. Several studies have shown the diagnostic accuracy of MR imaging in the detection and characterization of focal liver lesions in general (5 11). To our knowledge neither the incidence of benign liver lesions in patients with breast cancer with suspected liver metastases nor the diagnostic accuracy of MR imaging in the evaluation of the liver in this specific patient population have been described. The purpose of this study was to evaluate the MR findings in patients with known breast carcinoma and suspected liver metastases at initial investigation to determine the relative incidences of benign and malignant liver lesions. MATERIALS AND METHODS Patients The study included 34 patients (34 women; age range, years; mean age, 52 years) at our institution with known primary breast malignancies who had undergone MR imaging as part of their pre-treatment evaluation because of suspected liver metastases during a 5 year period between April 1993 and May Patients were referred for MR imaging to characterize further liver lesions demonstrated by computed tomography (CT) (n 9) or ultrasound (US) and CT (n 2) images, or for initial investigation without prior imaging studies because of liver enzyme abnormalities (n 13). No other specific selection or exclusion criteria were employed. All patients had histopathologic confirmation of primary breast malignancy. Twenty-four patients had infiltrating ductal carcinoma, one patient had ductal carcinoma in situ, five patients had invasive lobular

2 166 Noone et al. carcinoma, two patients had biphasic infiltrating ductal and invasive lobular carcinoma, one patient had biphasic ductal carcinoma in situ and invasive lobular carcinoma, and one patient had metaplastic carcinoma with squamous cell and infiltrating ductal components. One patient had concomitant hepatocellular carcinoma, and one patient had concomitant metastatic carcinoid. MR Imaging MR studies were obtained on 1.5 T MR imagers (SP4000 or VISION; Siemens, Iselin, NJ). MR sequences included precontrast T1-weighted breath-hold spoiled gradientecho (SGE) (TR/TE /4, flip angle 80, 8 10 mm slice thickness, sections acquired in a 20 second breath-hold) and T2-weighted fat-suppressed spinecho or turbo (fast) spin-echo (TR/TE / msec). SGE imaging was repeated immediately and 45 seconds following gadopentatate dimeglumine (Magnevist; Berlex, Wayne, NJ) administration in a dosage of 0.1 mmol/kg. T1-weighted fat-suppressed spin-echo (TR/TE 500/14 msec) or fat-suppressed SGE (TR/TE /4 msec) were acquired beginning at 90 seconds following contrast administration. SGE was performed 5 10 minutes following contrast administration in all patients. Image Interpretation Prospective image interpretation was performed on all MR studies by an experienced radiologist (R.C.S.), as the official clinical interpretation. Retrospective review was performed by a second experienced radiologist (T.C.N.), blinded to the interpretations of the prospective reader. The clinical interpretation was used to determine the incidence of both benign and malignant liver lesions in patients with primary breast carcinoma who were investigated by MR imaging for suspected liver metastases. The interpreters were aware of the clinical history and histologic confirmation of primary breast carcinoma previously obtained in all patients. The prospective and retrospective readings were in complete agreement regarding detection and characterization of lesions. The final MR imaging diagnosis then was correlated with surgical, laboratory, and histopathological information, as well as clinical follow-up, obtained by subsequent chart review performed months (mean, 22 months) following the MR study. True positive, true negative, false positive, and false negative detection, as well as sensitivity and specificity were determined. RESULTS Thirty-four patients with newly diagnosed breast carcinoma were evaluated with MR imaging. Liver lesions were detected in 32 patients, in whom benign lesions were observed in 13, malignant lesions in 21, and no lesions in 2. Benign and malignant liver lesions coexisted in two of the above-described patients. Correct detection of presence of lesions and characterization of lesions occurred in 33 patients, and 1 patient had false-negative lesion detection. Among the 13 patients (38%) with benign liver lesions, 2 (6%) patients had both benign and malignant liver lesions, and 11 (32%) patients had benign lesions only. Seven of the 34 (21%) patients had lesions with MR imaging characteristics consistent with hemangiomas (Fig. 1), 5 (15%) had lesions consistent with simple hepatic cysts, 2 (6%) had lesions with signal characteristics consistent with adenomas (Fig. 2), 1 (3%) had focal fatty infiltration, and 1 (3%) patient had focal fatty sparing. Three (9%) patients had more than one type of benign lesion. No focal liver lesions were detected in two patients (6%). There were no errors in determination of benignancy or malignancy or failure to detect benign lesions in patients reported to have no liver lesions, as determined by histologic findings in five patients and clinical/imaging follow-up in the remainder of patients. MR imaging revealed a total of 21 patients (62%) with malignant liver lesions: 19 (56%) patients had breast cancer metastases (Fig. 3) (2 had coexistent benign lesions), 1 had carcinoid metastases, and 1 had hepatocellular carcinoma. Lesions interpreted as breast carcinoma metastases demonstrated decreased signal intensity or isointensity on T1-weighted images, variably increased signal intensity on T2-weighted images, and homogeneous, heterogeneous, or rim enhancement following gadolinium administration. The number of lesions ranged from one to too numerous to count. The maximum dimensions of the metastases ranged from 0.2 to 13.0 cm (mean, 1.5 cm). Most liver lesions in the study measured less than 2 cm. Histopathologic correlation of liver lesions was available in five patients. Three of these patients had metastatic breast carcinoma, one patient had carcinoid metastases, and one patient had hepatocellular carcinoma. A differential diagnosis of both hepatocellular carcinoma and breast cancer metastasis was provided for the solitary lesion in the patient who subsequently underwent biopsy confirming the presence of hepatocellular carcinoma. Focal liver lesions were not identified in one patient with proven peri-portal breast carcinoma metastases; prominent liver capsular enhancement was noted, however, on the MR examination. The image quality was diminished on the SGE sequences due to the patient s inability to suspend respiration for the entire sequence duration. These lesions also were not detected on ultrasound or spiral CT. The interpretations in the remainder of patients were correlated with clinical, laboratory, and imaging follow-up (n 29), occurring 6 62 months following MR examination. Overall, true positive detection of malignant liver lesions was 20/21, true negative was 13/13, false positive was 0/13, and false negative was 1/21 for a sensitivity of 95%, specificity of 100%, positive predictive value of 100%, and negative predictive value of 92%. DISCUSSION The results of this study demonstrate the common occurrence of benign liver lesions in patients with breast carcinoma investigated for suspected liver metas-

3 Benign Liver Lesions in Breast Cancer 167 Figure 1. Hemangioma in a 49-year-old woman with infiltrating ductal carcinoma. a: SGE image (TR/TE 140/4 msec, flip angle 80 ). b: T2-weighted fat-suppressed echo train SE image (TR/TE 4500/90 msec). c: Immediate post-gadolinium SGE image. d: 90 second post-gadolinium fat-suppressed SGE image (TR/TE 170/4 msec, flip angle 80 ). A lesion present in segment 4 of the liver is moderately low signal intensity on the T1-weighted image (a) and moderately high signal on the T2-weighted image (b). The lesion demonstrates peripheral nodular enhancement immediately follow gadolinium enhancement (c). The nodules coalesce with centripedal contrast enhancement on the more delayed image (d). The appearance of this lesion is diagnostic for hemangioma. tases at initial presentation. One-third of the patients in our study who were investigated for suspected liver metastases had only benign liver lesions. The prevalence of small benign liver lesions in patients investigated for cancer has been described in larger patient series using CT examinations (12,13). Neither study, however, addressed either the overall prevalence of benign liver lesions or the prospective performance of CT in describing the small lesions as benign, although it appears that all lesions were prospectively considered indeterminate because of their small size. Inability to characterize these lesions correctly may lead to unnecessary duplication of imaging studies and/or inappropriate limitation of therapeutic options. Conversely, failure to detect small metastases may result in inappropriate allocation to therapy, which may adversely affect patients overall healthcare. This study also reaffirms the high diagnostic accuracy of MR imaging in the detection and characterization of focal liver lesions. A recent study, in which this MR imaging protocol was compared with portal venous phase spiral CT in 89 patients, demonstrated an increased detection rate of 49.4% for MR imaging compared with CT and an increased characterization rate of 75.3% for MR imaging (5). The high diagnostic accuracy of MR imaging reported in this study was also associated with a significant positive impact on patient management. Many diagnostic and therapeutic protocols employed in management of the breast cancer patient population currently do not recommend initial evaluation of the liver with MR imaging, possibly due in part to the perceived higher cost and varying levels of optimization of MR imaging. Initial baseline investigation with a modality with high accuracy in the evaluation of liver lesions may prevent unnecessary duplication of imaging studies, or, more importantly, errors in patient staging. More streamlined MR protocols and newer, more reproducible imaging sequences should provide more practical means for routine baseline examination of these patients. The limitations of this study include the lack of histopathologic corrrelation available for most lesions

4 168 Noone et al. Figure 2. Hepatic adenoma in a 39-year-old woman with infiltrating ductal adenocarcinoma. a: SGE (TR/TE 140/4 msec, flip angle 80 ). b: T2-weighted fat-suppressed echo train SE (TR/TE 4500/95 msec). c: Out-of-phase SGE. d: Immediate post-gadolinium SGE (TR/TE 140/4 msec, flip angle 80 ). e: 90 second post-gadolinium SGE (TR/TE 140/4 msec, flip angle 80 ) (e) images. No liver lesions is appreciated on the T1-weighted (a) or T2-weighted (b) images. On the out-of-phase image (c) uniform loss of signal intensity is noted ofa2cmmass in segment 6 of the liver (arrow, c). The lesion enhances uniformly and greater than background liver on the immediate post-gadolinium image (arrow, d), and subsequently fades rapidly to near isointensity with liver (arrow, e). detected by MR imaging. This may reflect the high confidence level of referring clinicians at our institution for the accuracy of liver lesion evaluation provided by MR imaging. It is currently acceptable practice to follow suspected liver cysts and hemangiomas with imaging studies, rather than biopsy (8). Follow-up imaging of the patients in this study with MR-characterized benign lesions demonstrated no misdiagnoses, although one patient with no liver lesions visualized was found to have periportal metastases at biopsy. It should be noted that image quality of the MRI study in the latter patient was reduced by the inability of the patient to suspend respiration adequately for the SGE sequences or to breathe in a regular fashion for the breathing-averaged T2-weighted sequence. Management of this patient was not, however, substantially adversely affected, since the referring oncologist had a high index of suspicion for liver metastases, due to abnormal laboratory values,

5 Benign Liver Lesions in Breast Cancer 169 In summary, this study demonstrates the relatively high incidence of benign liver lesions in patients with breast cancer who have suspected liver metastases at initial presentation. We have also shown that MR imaging has a high diagnostic accuracy in the evaluation of these lesions. These findings suggest the usefulness of MR imaging for the investigation of patients with breast carcinoma and suspected liver lesions. Figure 3. Liver metastases in a 56-year-old woman with infiltrating ductal carcinoma. Immediate post-gadolinium SGE (TR/TE 140/4 msec, flip angle 80 ) image. The metastasis is well shown as an intense ring-enhancing lesion (arrow). and was aware of the poor quality of the MRI study. Appropriate biopsy was performed, and this was positive for periportal malignancy. Another limitation of this study was the relatively small patient population. This reflected the fact that this study was performed at a single, medium-sized institution. It also reflected the fact that most patients were referred to MR imaging for characterization of lesions previously detected, but not characterized, by either CT or US, or with liver enzyme abnormalities, therefore representing a small, selected patient population. The incidence of benign liver lesions in our patient population may be artificially higher, as patients with obvious metastases on CT may not have been referred to MRI. An offsetting effect may have occurred in that patients with obvious benign lesions on CT may also not have been referred. Our study does not control for these possibilities, as we only evaluated MR studies and only MR studies in patients with newly diagnosed breast cancer and suspected liver metastases. Our data also do not reflect the prevalence of benign liver lesions in all patients with breast cancer at initial presentation, which may in fact represent more clinically useful information. This again reflects our selection of only those patients referred to MRI with suspected liver metastases. REFERENCES 1. Feigelson HS, Henderson BE, Pike MC. Re: recent trends in U.S. breast cancer incidence, survival, and mortality rate. Comment on: J Natl Cancer Inst 1996;88: J Natl Cancer Inst 1997;89: Wingo PA, Ries LA, Rosenberg HM, Miller DS, Edwards BK. Cancer incidence and mortality, Cancer 1998;82: Fumoleau P. Treatment of patients with liver metastases. Anti- Cancer Drugs 1996;7(Suppl 2): Pontiggia P, Curto FC, Sabato A, Rotella GB, Alonso K. Is metastatic breast cancer, refractory to usual therapy, curable? Biomed Pharmacother 1995;49: Semelka RC, Worawattanakul S, Kelekis NL, et al. Liver lesion detection, characterization, and effect on patient management: comparison of single-phase spiral CT and current MR techniques. J Magn Reson Imaging 1997;7: De Lange EE, Mugler JP, Gay SB, et al. Focal liver disease: comparison of breath-hold T1-weighted MP-GRE MR imaging and contrast enhanced CT lesion detection, localization, and characterization. Radiology 1996;200: Larson RE, Semelka RC, Bagley AS, et al. Hypervascular malignant liver lesions: comparison of various MR imaging pulse sequences and dynamic CT. Radiology 1994;192: Semelka RC, Shoenut JP, Kroeker MA, et al. Focal liver disease: comparison of dynamic contrast-enhanced CT and T2-weighted fat-suppressed, FLASH, and dynamic gadolinium-enhanced MR imaging at 1.5 T. Radiology 1992;184: Stark DD, Wittenberg J, Butch RJ, Ferrucci JT Jr. Hepatic metastases: randomized, controlled comparison of detection with MR imaging and CT. Radiology 1987;165: Semelka RC, Shoenut JP, Ascher SM, et al. Solitary hepatic metastasis: comparison of dynamic contrast-enhanced CT and MR imaging with fast-suppressed T2-weighted, breath-hold T1-weighted FLASH, and dynamic gadolinium-enhanced FLASH sequences. J Magn Reson Imaging 1994;4: Yamashita Y, Mitsuzaki K, Yi T, et al. Small hepatocellular carcinoma in patients with chronic liver damage: prospective comparison of detection with dynamic MR imaging and helical CT of the whole liver. Radiology 1996;200: Jones EC, Chezmar JL, Nelson RC, Bernardino ME. The frequency and significance of small hepatic lesions ( 1.5 mm) detected by CT. AJR 1992;158: Schwartz LH, Gandras EJ, Colangelo SM, Ercolani MC, Panicek DM. Prevalence and importance of small hepatic lesions found at CT in patients with cancer. Radiology 1999;210:71 74