Early diagnosis of recurrent breast cancer is crucial to selection of the most appropriate therapy. The current study evaluated the role of FDG-PET/CT in the assessment of suspected recurrent breast cancer in patients who presented with elevated serum tumor markers.
Forty-seven consecutive FDG-PET/CT studies of 46 women (aged 32–79 years; mean, 59.9 years) with a history of breast cancer presented with elevated serum tumor markers 1–21 years (mean = 6.2 years) after their initial diagnosis and were retrospectively evaluated. PET/CT results were confirmed by pathology (n = 11), further imaging, and follow-up (mean = 17.2 months; n = 36). Changes in further management based on PET/CT were recorded.
Thirty (65%) patients had tumor recurrence, and 16 (35%) patients showed no further evidence of disease. Thirty-one patients had 32 abnormal PET/CT studies, and 15 patients had normal studies with an overall sensitivity, specificity, and accuracy of 90%, 71%, and 83%, respectively. In 37 patients, PET/CT was compared with contrast-enhanced CT and had a higher sensitivity (85% vs 70%), specificity (76% vs 47%), and accuracy (81% vs 59%). PET/CT had an impact on the management of 24 (5l%) patients. Of these, chemotherapy or radiotherapy was started in 16 patients, treatment was modified in 2 patients, and 6 patients were referred to biopsy, followed by referral to surgery for 2 patients.
Early diagnosis and accurate restaging of recurrent breast cancer is important to the selection of the most appropriate therapeutic strategy, mainly by identifying patients with limited disease who could benefit from curative treatment. Positron emission tomography with 18fluorodeoxyglucose (FDG-PET) has been used for diagnosis, staging, monitoring response to therapy, and restaging patients with breast cancer.1–6 Although FDG-PET may have limited diagnostic value in detecting small primary breast tumors, in staging of the axillary region, and in detecting osteoblastic metastases, it is superior to conventional imaging modalities in detecting distant metastases and in monitoring response to therapy.1 Whole-body FDG-PET is of clinical value in the search for metastases, mainly when suggested by the presence of clinical symptoms or by a progressive increase in biochemical markers (CA15-3, CEA).4, 7–10 In asymptomatic breast cancer patients with rising serum tumor markers, metastatic disease can be diagnosed by FDG-PET with an accuracy of 87% to 90%, compared with 50% to 78% for conventional imaging.8, 9, 11
Integrated PET/computed tomography (PET/CT), which combines anatomic and metabolic imaging information, has been shown to further improve diagnostic accuracy and clinical management of patients through accurate localization of functional data on high-resolution anatomic CT images.11, 12
The purposes of the current study were to evaluate the accuracy of FDG-PET/CT for detection and restaging of recurrent breast cancer and the impact of this imaging modality on the management of patients who present with elevated serum tumor markers.
MATERIALS AND METHODS
Forty-seven consecutive FDG PET/CT studies of 46 women (aged 32–79 years; mean, 59.9 years) with elevated serum tumor markers, and who were assessed for suspected recurrence of breast cancer, were retrospectively evaluated. Thirty-four patients presented with rising CA-15-3, 9 patients had increased CEA, 5 patients had elevated CA-125, and 3 patients had elevated CA 19-9. Five patients had more than 1 elevated serum tumor marker (Table 1). Contrast-enhanced high-resolution CT was performed in 37 patients 1–4 months (median = 2.1 months) before PET/CT was performed.
Table 1. Clinical Characteristics of 46 Patients With Suspected Recurrent Breast Cancer
Time from last treatment, y
Treatment no. patients (%)
Surgery, chemo- & radiotherapy
Surgery & chemotherapy
Surgery & radiotherapy
Hormonal therapy (Tamoxifen)
Elevated tumor markers no. patients (%)
The time interval between initial diagnosis of breast cancer and the FDG-PET/CT study was 1–21 years (mean, 6.2 years). The time interval after the last treatment was 0.3–15 years (mean, 4.6 years). One patient had rising tumor markers during treatment (Table 1). Prior treatment included surgery, chemotherapy, and radiotherapy in 29 patients; surgery was followed by chemotherapy in 6 patients; surgery and radiotherapy was given to 6 patients; surgery only was performed on 5 patients, and chemotherapy was given to only 1 patient. Treatment with hormones had been previously administered to 14 patients (Table 1).
The patients were studied using a dedicated PET/CT system (Discovery LS; General Electric Healthcare Technologies, Milwaukee, WI). Patients were instructed to fast for 4–6 hours before injection of 18F-FDG except for glucose-free oral hydration. Blood glucose was measured before injection of a tracer to ensure a level of <11 mmol/L. The injected dose of 18F-FDG was 370–666 MBq. After injection, patients were kept lying comfortably. The whole-body FDG-PET was performed 60 minutes after 18F-FDG injection.
The PET/CT system comprises a dedicated PET scanner with a full-ring bismuth germanate detector and a multislice CT. The protocol included an initial CT acquisition followed by PET study. CT parameters used for acquisition included 140 kV, 80 mA, 4-slices helical, 0.5 seconds per rotation, and pitch of 6:1, with a slice thickness of 4.25 mm, equal to that of PET. CT images were reconstructed to a 512 × 512 matrix. PET was acquired by sequential fields of view, each covering 15 cm during an acquisition time of 4 minutes. PET acquisition was performed in 2-dimensions with a matrix of 128 × 128. PET data were reconstructed by using order-subsets expectation maximization. Data obtained from CT acquisitions were used for low-noise attenuation correction of PET emission data and for fusion of attenuation-corrected PET images with corresponding CT images.
After completion of PET acquisition, the reconstructed x-ray attenuation-corrected PET images, CT images, and fused images were available for review in axial, coronal and sagittal planes, and in maximum-intensity-projection 3-dimensional cine mode, using the manufacture's review station (Xeleris; General Electric Healthcare Technologies, Milwaukee, WI).
A team of experienced nuclear medicine physicians and body radiologists interpreted the 18F-FDG PET/CT images with knowledge of clinical history and previous imaging study results (current authors L.R., R.B-S., L.G., O.I.). The presence and localization of any area of increased 18F-FDG uptake were recorded on PET/CT, and each lesion was characterized as benign, malignant, or equivocal. A focus of increased 18F-FDG was defined as benign when related to the physiologic biodistribution of 18F-FDG or to a known nonmalignant process. A focally abnormal 18F-FDG activity of higher intensity than that of surrounding tissues, which could not be related to benign or physiologic 18F-FDG uptake or was localized in a suspicious mass on CT, was defined as malignant. Any area of increased 18F-FDG uptake that could not be clearly characterized was defined as equivocal. Disagreements on final interpretation were resolved by a majority opinion of the combined team (current authors L.R., R.B-S., L.G., O.I.).
For the site-based analysis a true-positive lesion was defined as malignant or equivocal on PET/CT with subsequent confirmed tumor involvement. A false-positive site was defined as malignant or equivocal on PET/CT with no further evidence of disease. A true-negative site was defined as benign or physiologic on PET/CT with no further evidence of disease. A false-negative site was defined as benign or physiologic on PET/CT showing subsequent evidence of malignancy.
For the patient-based analysis, a PET/CT study was defined as a true positive when it showed at least 1 malignant or equivocal focus of FDG uptake with further confirmed malignancy. A false-positive study showed foci of FDG uptake defined as malignant or equivocal with no evidence of active cancer in further evaluation. A study that showed only benign, or no abnormal findings, was defined as a true negative when the patient had no further evidence of active cancer, and a study was defined as a false-negative when the patient had active disease on follow-up.
The reports of previous diagnostic CT scans were used for comparison with PET/CT studies. A CT study with findings suspicious for malignancy was defined as a true positive when malignancy was proven and as a false-positive when there was no evidence of malignant disease during follow-up. A normal CT or CT showing only benign findings in patients with no further evidence of malignancy was defined as a true negative, and as a false-negative when malignancy was proved during follow-up. The final diagnosis was obtained by histopathology (n = 11) or by combined follow-up (n = 36), including imaging studies (ultrasound, mammography and/or repeat CT or PET/CT) and clinical reassessment (n = 36; 1–42 months; median, 17.2 months). The clinical follow-up period for patients with no evidence of disease was 10 to 33 months (median, 21.9 months).
The impact of PET/CT studies on the management of patients was evaluated on the basis of clinical decisions obtained from patient files and interviews of referring physicians. Intramodality or intermodality changes in treatment strategies, as well as guidance for further diagnostic procedures (ie, biopsy), were recorded.
Sensitivity, specificity, accuracy, positive predictive value and negative predictive value of PET/CT and high resolution, contrast-enhanced CT for diagnosis of recurrent breast cancer were calculated for both site-based and patient-based analysis.
Recurrence of breast cancer was diagnosed in 30 (65%) patients, whereas 16 patients with 17 (35%) studies had no further evidence of active malignancy.
One hundred seventy-one sites of increased FDG uptake were detected in 37 of 47 of the PET/CT studies (Table 2). Ten studies showed no abnormal foci of FDG uptake. Malignancy was diagnosed in 153 of the 171 (89%) lesions. The sites of increased FDG uptake were localized in the neck (n = 18), thorax (n = 88), abdomen (n = 32), pelvis (n = 24), and thighs (n = 9). PET/CT enabled correct localization of 68 (40%) lesions in 23 patients, including 9 of 18 (50%) sites in the neck, 46 of 88 (52%) sites in the thorax, 7 of 32 (22%) sites in the abdomen, and 6 of 24 (25%) sites in the pelvis. PET/CT characterized 108 foci as malignant, 11 as benign, and 52 as equivocal. There were 151 true-positive lesions, 5 false-positive, 13 true-negative, and 2 false-negative lesions. The site-based performance characteristics of FDG-PET/CT are detailed in Table 2.
Table 2. Number and Location of 171 FDG-Avid Sites
There were 32 abnormal PET/CT studies in 31 patients, 27 were true-positive, and 5 were false-positive. Twenty-four of the 27 true-positive studies showed disseminated disease, including 5 patients with multiple bone metastases (Fig. 1), 4 patients with nodal disease, 4 patients with bone and lymph node involvement, 3 patients with multiple liver lesions, and 8 patients with lymph node, bone and liver, or lung metastases. Three of the 27 true-positive studies showed a single metastatic focus (Fig. 2).
There were 5 false-positive studies in 4 patients. Two studies in 1 patient showed increased FDG uptake in lung lesions seen on CT. These pulmonary lesions disappeared with no therapy on a repeat CT study performed 9 months after the second PET/CT study, and tumor markers returned to normal values. These findings were, therefore, considered to represent an inflammatory process. The second patient showed increased FDG uptake in residual left breast tissue 2 years after left mastectomy. Mammography was negative, and a biopsy performed 6 months later showed no evidence of malignancy. In a third patient, PET/CT demonstrated increased FDG uptake in an enlarged left cervical lymph node. Repeat measurements of serum tumor markers after 1 month were normal, and there was no evidence of malignancy during 33 months follow-up. The FDG-avid cervical lymph node was, therefore, considered to be reactive to inflammation. The fourth patient showed a suspicious focus in the thyroid. Ultrasound of the neck diagnosed a multinodular goiter. There was no evidence of malignancy during 11 months follow-up.
Fifteen patients had 15 negative FDG-PET/CT studies. Twelve (80%) patients showed no further evidence of malignancy for 10–40 months follow-up (median, 23.3 months) and were defined as true negative.
Three PET/CT studies were false-negative, with malignancy diagnosed at 1–6 months after the study. One patient showed increased FDG uptake in her healthy breast, considered to represent physiologic tracer activity. One month later, a lesion in the breast, 7 mm in diameter, was detected on mammography; adenocarcinoma was diagnosed on fine needle aspiration, and the patient underwent lumpectomy and chemotherapy. A second patient developed an intestinal obstruction 2 months after a negative finding by a PET/CT study. Multiple peritoneal metastases from breast carcinoma were diagnosed at surgery. Squamous cell carcinoma, most probably metastatic from the breast, was diagnosed in the third patient in a 9 mm pulmonary lesion at 5 months after a negative finding by PET/CT.
The patient-based performance characteristics of PET/CT are detailed in Table 3.
Table 3. Performance Indices of PET/CT for Diagnosis of Recurrent Breast Cancer in 47 Studies and 171 Lesions
High-resolution contrast-enhanced CT studies were available for comparison with PET/CT in 37 patients. CT was positive and raised the suspicion of recurrent malignancy in 23 patients. In 14 of these patients, malignancy was further confirmed by a true-positive result. PET/CT was positive in 13 and negative in 1 patient in this group. Nine patients with abnormal CT had no further evidence of malignancy (false positive). PET/CT was negative in all of these patients.
Fourteen patients had negative findings by CT studies. Eight of these patients had no further evidence of malignancy (true negative). PET/CT was positive in 4 and negative in 4 patients in this group. In 6 patients with negative CT findings, recurrent disease was diagnosed on follow-up (false positive). PET/CT was positive in 4 and negative in 2 of these patients (Table 4).
Table 4. Comparison of Contrast-enhanced CT and PET/CT Performance in 37 Patients With Suspected Breast Cancer Recurrence
Among the subgroup of 37 patients, CT and PET/CT showed congruent results in 19 patients. PET/CT was the single modality for diagnosis of recurrent malignancy in 4 patients, and CT diagnosed recurrent cancer in 1 patient. In addition, PET/CT demonstrated more extensive disease compared with contrast-enhanced CT in 13 patients.
PET/CT Impact on Patient Management
PET/CT results changed the management of 24 (51%) patients. Chemotherapy or radiotherapy was started in 16 patients after abnormal PET/CT studies showed previously unknown disease. Treatment was changed in 2 patients with more extensive disease than initially suspected by CT. In 1 patient, radiofrequency ablation of a liver metastasis was cancelled, and chemotherapy was started. In the second patient, hormonal therapy was discontinued, and she was started on chemotherapy. PET/CT-guided biopsy for tissue sampling was performed in 6 patients (3 lung lesions, 1 thyroid nodule, 1 breast nodule, 1 bowel lesion), and 2 of these patients were further referred to surgery (pneumonectomy and hemicolectomy).
Recurrent breast cancer can pose considerable diagnostic and therapeutic challenges for the oncologic team. Prognostic and management decisions are based on early knowledge of the whole extent of disease.1–4 Conventional restaging methods may not reliably demonstrate the extent of disease. On CT, sequelae of previous surgery and radiation therapy can be difficult to distinguish from recurrent neoplasms, and small foci of early metastatic disease can be difficult to characterize and differentiate from benign lesions.10
FDG-PET is accurate in restaging recurrent breast cancer, by clarifying inconclusive findings on physical examination and on cross-sectional anatomic imaging, and has been proven accurate in restaging recurrent breast cancer.13 A major advantage of FDG-PET compared with conventional imaging is the screening of the whole body. For detection of recurrent breast cancer, FDG-PET has a reported average sensitivity and specificity of 96% and 77%, respectively, and, as a rule, the sensitivity of FDG-PET is higher than that of a combination of conventional imaging methods.7, 8 By a systematic search of articles and meta-analysis of articles published in medical journals, Isasi et al.14 evaluated the diagnostic performance of FDG-PET in detecting breast cancer recurrence and metastases. Among the studies with patient-based data, sensitivity ranged from 55.6% to 100% (median, 92.7%), and specificity ranged from 0% to 100% (median, 81.6%). In studies with lesion-based data, sensitivity ranged from 57% to 97% (median, 91.7%), and specificity ranged from 79% to 96% (median, 88.9%). Although FDG-PET has good sensitivity for diagnosis of breast cancer recurrence and metastases, its use was limited in the follow-up of breast cancer patients by a relatively high rate of false-positive results, which were caused by increased glucose utilization in benign tumors and in inflammatory or infectious processes.
Only few studies have previously assessed the role of integrated PET/CT in patients with breast cancer, and those studies were of nonhomogeneous populations. The specificity of FDG-PET has been shown to improve significantly by using the combined anatomical-molecular PET/CT imaging technique.9, 12 In a retrospective review of 75 patients with suspected breast cancer, Tatsumi et al compared performance of PET, CT, and PET/CT.9 PET/CT resulted in improved diagnostic confidence compared with PET in 60% of patients and in 55% of regions and resulted in significantly better accuracy compared with CT (P < .05). Fueger et al reported improved sensitivity and specificity of PET/CT compared with PET (94% and 84% compared with 85% and 72%, respectively) in 58 patients with suspected recurrent breast cancer.12 The specificity of FDG-PET was only partially improved by the use of PET/CT because of FDG-avidity of falsely positive inflammatory lesions despite the use of PET/CT.
In the current study, the sensitivity of FDG-PET/CT for diagnosis of recurrent tumor in breast cancer patients with elevated serum tumor markers was high, at 90%. False-negative findings were due to small lesions and peritoneal metastases. The 5 false-positive studies, all with high FDG uptake, were due to inflammatory or other benign processes and resulted in a specificity of 71%. In comparison with CT, integrated PET/CT imaging showed improved sensitivity, specificity, accuracy, positive predictive value, and negative predictive value. Most importantly, the current study demonstrates that FDG-PET/CT results had an impact on the management of 51% of patients.
The main limitation of the current study, its retrospective nature that might have introduced a referral bias, is partly overcome by use of consecutive patient data. Also, not all suspected lesions had confirmed pathology laboratory results. Routine criteria, including repeat imaging studies and prolonged clinical follow-up, were used instead to determine the nature of the findings, similarly to previous literature reports. Larger prospective patient studies are needed to confirm these initial results concerning the added role of FDG-PET/CT for patient care in suspected recurrent breast cancer that presents with elevated serum tumor markers.
Integrated FDG-PET/CT is a sensitive and accurate imaging modality, superior to CT for diagnosis of tumor recurrence and for definition of extent of disease in patients with breast cancer and rising tumor markers. PET/CT appears to have a role in determining the subsequent clinical management of these patients.