Nuclear Medicine and Molecular Imaging • Original Research Quartuccio et al. PET/CT in Pediatric Bone Sarcoma
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Nuclear Medicine and Molecular Imaging Original Research
Pediatric Bone Sarcoma: Diagnostic Performance of 18F-FDG PET/CT Versus Conventional Imaging for Initial Staging and Follow-Up Natale Quartuccio1 Josef Fox 2 Deborah Kuk 3 Leonard H. Wexler4 Sergio Baldari1 Angelina Cistaro5,6,7 Heiko Schöder 2 Quartuccio N, Fox J, Kuk D, et al. Keywords: 18F-FDG PET/CT, Ewing sarcoma, osteosarcoma, pediatric bone sarcoma, prognosis DOI:10.2214/AJR.14.12932 Received March 28, 2014; accepted after revision July 3, 2014. N. Quartuccio was a recipient of the 2011 Bradley-Alavi Student Fellowship Award. N. Quartuccio and A. Cistaro contributed equally to this work. 1
Nuclear Medicine Unit, Department of Biomedical Sciences and of Morphologic and Functional Images, University of Messina, Messina, Italy.
2
Molecular Imaging and Therapy Service, Department of Radiology, Memorial Sloan Kettering Cancer Center, 1275 York Ave, New York, NY 10061. Address correspondence to H. Schöder ([email protected]). 3 Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY. 4 Department of Pediatrics, Memorial Sloan Kettering Cancer Center, New York, NY. 5
PET Centre IRMET S.p.A., Euromedic, Turin, Italy.
6
Coordinator of PET Pediatric AIMN InterGroup, Turin, Italy.
7
Institute of Cognitive Sciences and Technologies, National Research Council, Rome, Italy.
AJR 2015; 204:153–160 0361–803X/15/2041–153 © American Roentgen Ray Society
OBJECTIVE. The purpose of this study was to compare the diagnostic performance of PET/CT and conventional imaging for staging and follow-up of pediatric osteosarcoma and skeletal Ewing sarcoma. MATERIALS AND METHODS. We calculated sensitivity, specificity, and accuracy of PET/CT and conventional imaging (CT, MRI, bone scanning) for sites of disease and number of lesions. Diagnostic benefit, defined as better characterization of lesions, was evaluated on a per-scan basis, comparing PET/CT and conventional imaging. RESULTS. A total of 412 lesions were characterized by imaging in 64 patients (20, osteosarcoma; 44, Ewing sarcoma). For osteosarcoma patients PET/CT was available only at follow-up, where it proved more accurate than conventional imaging for the detection of bone lesions (accuracy, 95% vs 67% for CT and 86% for MRI) and complementary to CT in evaluating lung nodules (sensitivity, 84% vs 94%; specificity, 79% vs 71%) with diagnostic benefit in 18% of examinations. In patients with Ewing sarcoma, PET/CT tended to perform better during follow-up than at initial staging (accuracy, 85% vs 69%). For lung findings, PET/CT was more specific than CT but was less sensitive. The diagnostic benefit of PET/CT was greater at staging (28%) than during followup (9%). On a per-patient basis, PET/CT provided diagnostic benefit in 21 of 44 patients with Ewing sarcoma and nine of 20 patients with osteosarcoma at least once during clinical management. CONCLUSION. FDG PET/CT provides diagnostic benefit in Ewing sarcoma and osteosarcoma, with the exception of small lung nodules. Prospective studies are needed to define the best imaging algorithm and combination of tests in the staging and follow-up of patients with pediatric bone sarcoma. 18F-FDG
T
he most common primary malignant bone tumors in children are osteosarcoma and Ewing sarcoma [1]. Osteosarcoma derives from primitive bone-forming mesenchymal stem cells and occurs mainly in the metaphyseal portion of the long bones [2]. Ewing sarcoma is equally divided between the axial and the appendicular skeleton [3, 4]. Both tumors have a peak incidence in the second decade of life, and the presence of metastases at diagnosis is one of the strongest predictors of survival [2, 5]. Approximately 20–25% of patients with osteosarcoma and Ewing sarcoma present with radiographically detectable distant metastases [6, 7]; however, the proportion of patients with lung-only metastases is substantially greater in osteosarcoma than in Ewing sarcoma, in which bone or bone marrow metastases are seen alone or in combination with lung metastases in nearly one half of patients with distant metastases
[8, 9]. Lymph node and visceral metastases are uncommon in both tumor types at diagnosis but may be seen in the setting of advanced, recurrent disease [6, 10]. The 5-year survival for both malignancies is 65–70% with localized disease compared with only 20% with detectable metastatic disease [2, 6]. Accurate detection of metastases is important for risk stratification and may influence the clinical management of these patients. An important imaging modality for assessing bone sarcomas is 18F-FDG PET/CT [11]. Few studies have investigated the diagnostic performance of FDG PET/CT in children with osteosarcoma and Ewing sarcoma [8–11]. Presently, FDG PET/CT is not considered a standard examination in the diagnostic algorithm for these patients [12, 13]. Nevertheless, in our clinical experience FDG PET/CT accurately depicts metastatic lesions from bone sarcoma both at diagnosis and at the time of relapse. The purpose of
AJR:204, January 2015 153
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al. this study was to compare the performance of FDG PET/CT with conventional imaging for detecting osteosarcoma and Ewing sarcoma lesions in different anatomic regions. Additionally, we aimed to determine the potential impact of FDG PET/CT results on patient management. Materials and Methods Patients This retrospective study was approved by our institutional review board with a waiver of authorization. Patient consent was not required. A retrospective search of our institutional PET database from February 2002 to September 2012 was conducted using the following criteria: age more than 18 years at diagnosis; biopsy-proven diagnosis of osteosarcoma or skeletal Ewing sarcoma; availability of at least one FDG PET/CT study for tumor staging or follow-up at our institution; and availability of one or more conventional imaging modalities, including dedicated CT, MRI, or bone scanning performed within 1 month of the comparison PET/CT examination. For evaluation of pulmonary nodules, the availability of a dedicated chest CT study (with or without IV contrast material) was required as the comparator. Bone lesions were assessed using all imaging techniques. Lymph node metastases were evaluated by MRI and CT. Other sites (such as soft tissue, liver, and brain) were evaluated by the most relevant conventional imaging modality. Imaging datasets and reports of PET/CT and conventional imaging were reviewed. Histopathologic examination of lesions (if available) or clinical and imaging follow-up for at least 6 months were considered the standard of reference.
PET/CT Examination and Image Analysis PET/CT studies were performed on various systems (Discovery LS, Discovery ST, Discovery STE, or Discovery PET/CT 610, all GE Healthcare). Patients fasted for at least 6 hours before scanning. Plasma glucose tested before FDG injection was less than 160 mg/dL in all patients. PET/CT scans were acquired in 2D mode before 2008 and in 3D mode thereafter. The average uptake time was 72 minutes; injected activity ranged from 113 to 596 MBq, depending on the body surface area. Images were reconstructed using dedicated software. Scans were visually inspected for the presence of abnormal uptake distinguishable from background activity and suggestive of tumor metabolism. Maximum SUV values (SUVmax) normalized for body weight were then measured for each lesion [14, 15]. Reference background FDG activity was measured in the right hepatic lobe and
154
TABLE 1: Patient Characteristics Characteristics
All Patients
Ewing Sarcoma
Osteosarcoma
64
44
20
Median age at diagnosis (y)
13.2
13.3
13.1
Median age at PET/CT (y)
15.0
14.7
15.5 10
No. of patients
Female
34
24
Male
30
20
10
421
346
75
Total no. PET/CT No. of PET/CT at staging
18
18
0
No. of PET/CT at follow-up
403
328
75
Total no. of lesions found
412
311
101
Mean SUV liver
2.3
2.3
2.41
Mean SUV blood pool
1.6
1.6
1.7
Background regions
Primary lesions Mean SUVmax (g/L)
4.9 (1.9–10.8)
5.0 (1.9–10.8)
NA
Mean dimensions (mm)
49.9 (2.9–175)
52.7 (2.9–175)
NA
Mean SUVmax range (g/L)
2.3 (0.2–6.3)
0.8 (0.2–5.6)
2.9 (0.4–6.3)
Mean dimensions range (mm)
9 (1.7–59.7)
6.9 (1.7–48.8)
19 (2.0–59.7)
Lung lesions
Note—Data in parentheses are range. SUVmax = maximum SUV. NA = not applicable.
mediastinal blood pool [16]. Lesion size was measured on MRI or CT, depending on the location.
Statistical Analysis Patient and scanning characteristics are presented using the median, mean, and range for continuous variables and counts for categoric variables. FDG PET/CT studies were compared with CT, MRI and bone scanning studies in a lesionbased (pulmonary, bone, lymph node, or other) analysis at initial staging and follow-up and for each disease. According to the standard of reference, imaging findings were retrospectively classified as true-positive, true-negative, false-positive, or false-negative. Findings described in the reports as “equivocal,” “not conclusive,” or “suggestive of malignancy” were considered PETpositive and then classified as true-positive or false-positive lesions using the same standard of reference. Concordance between PET/CT and conventional imaging results was assessed. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated on a lesional and patient basis and categorized by time point (initial staging versus follow-up). Finally, we investigated whether PET/CT results provided a diagnostic benefit beyond that of conventional imaging. This was determined by comparing the number of true-positive and false-positive lesions recorded per scan; recording of more true-positive lesions or fewer
false-positive lesions on PET versus conventional imaging represented a diagnostic benefit. All statistical analyses were performed using R, version 3.0.2 (R Project). A p value of less than 0.05 was considered significant.
Results We retrieved 421 PET/CT examinations. A total of 412 lesions were compared on FDG PET/CT and conventional imaging in a total of 64 patients (Table 1). Pathology results were available for 67 of 412 findings (20, pulmonary; 47, nonpulmonary); a minimum 6 months of follow-up was used for confirmation of the remaining cases. This analysis includes 16 primary Ewing sarcomas; of the remaining 396 lesions, 271 were eventually classified as metastasis on the basis of the standard of reference. Ewing Sarcoma A total of 311 lesions (lung, 132; bone, 115; lymph node, 32; and other sites, 32) were characterized on imaging with an overall concordance of 79.2% (staging, 68.8%; follow-up, 80.9%) between FDG PET/CT and conventional imaging (Tables 2–4). Primary tumors (n = 16) showed variable FDG uptake, with SUVmax between 1.9 and 10.8 (Table 1). In the per-scan comparison, a diagnostic benefit was seen in 27.8% of initial
AJR:204, January 2015
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
staging PET/CT studies (Fig. 1), with fewer false-positive lesions compared with CT in two scan-sets and more true-positive lesions in three scan-sets. For follow-up scans, a diagnostic benefit was seen in 9.1% of cases (30/328 datasets), including fewer false-positive findings in 17 examination sets and more true-positive lesions in 13 examination sets. Lung lesions—At initial staging, all three malignant nodules were correctly classified on dedicated chest CT but not on PET/CT (lesion size, 6.1, 2.3, and 5.6 mm; SUV, 0.4, 0.5, and 0.5, respectively). At follow-up, 25 of the 28 malignant nodules were detected on chest CT (mean size, 6.9 mm; range, 1.7– 25.5 mm). The average SUV of these nodules was 1.0. Four of the 28 nodules were false-negative on PET/CT (size range, 1.7– 3.5 mm; all with SUVmax ≤ 0.2). In comparison with CT, PET/CT had a slightly lower sensitivity (85.7% vs 89.3%) but a slightly higher specificity (90.7% vs 84.9%). Representative cases are shown in Figures 2 and 3. Bone lesions—At initial staging, a total of 20 metastatic lesions, distant from the primary tumor, were identified. For lesions within the FOV of each imaging test, MRI showed sensitivity of 100% and the highest overall accuracy (88.9%). CT performed the worst, detecting only four of nine metastases. Whole-body PET/CT detected all 16 malignant lesions (sensitivity 100%) but at the expense of a low specificity of 25%. At follow-up, a total of 95 bone lesions (10 suspected local recurrences and 85 distant sites) were identified, of which 50 were malignant. In the overall assessment, PET/CT showed slightly higher sensitivity than did MRI, with similar specificity. Among the 10 suspected local recurrences (on the basis of clinical examination or findings in one of the imaging tests) only three were found to be true. These three lesions were detected by both PET and bone scanning, but the bone scanning study showed more false-positive sites. Two of the three lesions were also positive on MRI; the third lesion was not imaged with MRI. Among the eight local recurrence sites with available PET and MRI, imaging findings were concordant for six and discordant for two (false-positive that was true-negative or true-negative that was false-positive result). Lymph nodes—A total of 32 nodal lesions were identified either at initial staging (n = 4) or during follow-up (n = 28). At staging, all metastatic nodes were correctly identified by PET/CT or CT/MRI. At follow-up, eight FDGavid nodes proved to be false-positive (four in the mediastinum and one each in the subpectoral, axillary, cervical, and inguinal regions). MRI also showed two false-positive nodes.
TABLE 2: Ewing Sarcoma Overall Analysis: Lesion-Based Diagnostic PET/CT in Pediatric Bone Sarcoma Performance Combined for Staging and Follow-Up No. of Lesions
Sensitivity (%)
PET/CT
311
90.3
CT
231
71.1
MRI
95
92.1
77.2
Bone scanning
91
42.3
66.7
Examination
Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
77.8
69.9
93.3
82
84.5
69.2
85.6
80.1
72.9
93.6
83.2
62.9
46.4
52.8
Note—PPV = positive predictive value, NPV = negative predictive value.
TABLE 3: Ewing Sarcoma Staging: Diagnostic Performance by Disease Site Lesion
No. Sensitivity (%) Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
Lung PET/CT
18
0
53.3
0
72.7
44.4
CT
18
100
46.7
27.3
100
55.6
20
100
25
84.2
100
85
Bone PET/CT CT
9
44.4
Not a numbera
100
0
44.4
MRI
9
100
75
83.3
100
88.9
Bone scanning
20
62.5
100
100
40
70
4
100
66.7
50
100
75
CT
3
Not a numbera
100
Not a numbera
100
100
MRI
4
100
100
100
100
100
3
100
100
100
100
100
CT
2
Not a numbera
100
Not a numbera
100
100
MRI
1
100
Not a numbera
100
Not a numbera
100
Lymph node PET/CT
Other PET/CT
Note—PPV = positive predictive value, NPV = negative predictive value, NA = not assessable. aOccurs when the denominator is zero.
TABLE 4: Ewing Sarcoma Follow-Up: Diagnostic Performance by Disease Site Lesion
No. Sensitivity (%) Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
Lung PET/CT
114
85.7
90.7
75
95.1
89.5
CT
114
89.3
84.9
65.8
96.1
86.0 86.3
Bone PET/CT
95
96.0
75.6
81.4
94.4
CT
46
57.1
83.3
84.2
55.6
67.4
MRI
56
88.0
74.2
73.3
88.5
80.4
Bone scanning
71
33.3
62.9
48.0
47.8
47.9
Lymph node PET/CT
28
100
66.7
33.3
100
71.4
CT
22
75.0
100
100
94.7
95.6
MRI
7
Not a numbera
71.4
0
100
71.4
PET/CT
29
80.0
68.4
57.1
86.7
72.4
CT
17
75.0
100
100
92.9
94.1
MRI
18
100
83.3
75.0
100
88.9
Other
Note—PPV = positive predictive value, NPV = negative predictive value. aOccurs when the denominator is zero.
AJR:204, January 2015 155
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al.
A
B
Fig. 1—Example of diagnostic benefit of 18F-FDG PET at staging in 14-year-old girl with primary Ewing sarcoma of right femur. A, Axial low-dose CT (left) and fused PET/CT (right) images show extensive hypermetabolic mass involving right proximal femur and thigh. B, Coronal low-dose CT (left), fused PET/CT (center), and PET (right) images clearly depict extension of hypermetabolic tumor into right femoral vein (arrow). C, Maximum-intensity-projection (left) and fused axial PET/CT (right) images show additional FDGavid lesion within right ventricle of heart (arrows) suspicious for tumor thrombus in view of right femoral vein involvement. D, Axial T1 fat-saturated gadolinium-enhanced cardiac MR image obtained 1 day later confirmed mobile pedunculated mass (arrow) adjacent to free wall of right ventricle. At surgery, 3 × 2 × 1 cm mass was resected, yielding metastatic Ewing sarcoma. E, Intracardiac lesion was not detected prospectively on contrast-enhanced CT performed 1 week before PET/CT, although right ventricular filling defect is evident on retrospective review. In either case, FDGavidity on PET enabled diagnosis of tumor thrombus (as opposed to bland thrombus).
156
C
D
E AJR:204, January 2015
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
PET/CT in Pediatric Bone Sarcoma Other sites—There were a total of 32 abnormal findings in other body regions at initial staging (liver, n = 2; skeletal muscle, n = 1) and during follow-up (liver, n = 3; skeletal muscle, n = 21; and n = 1 each for brain, heart, scalp, nasopharynx, and pleura). At initial staging, there was no difference in diagnostic performance between PET/CT and conventional imaging; both showed an accuracy of 100%. During follow-up, MRI had the highest sensitivity (100%, n = 18) and CT had the highest specificity (100%, n = 17). On PET/CT, two lesions were false-negative, one in the paraspinous soft tissues and one in the left forearm. Osteosarcoma Only follow-up PET/CT examinations were available for osteosarcoma patients (Table 5). Concordance of imaging findings on PET/CT versus conventional imaging was noted in 82.4% of cases. Diagnostic benefit was provided by PET/CT in 17.4% of datasets, showing fewer false-positive lesions in 6.7% (5/75) and more true-positive lesions in 10.7% (8/75). Lung lesions—A total of 32 malignant and 24 benign nodules were detected. The mean size of the malignant nodules was 19 mm (range, 2–60 mm); the mean SUV was 2.9 (range, 0.4–6.3). Dedicated CT had a lower rate of false-negative findings than did PET/CT. The two nodules that were falsely reported as benign on CT measured 5 and 2 mm. PET/CT failed to correctly characterize five malignant lung nodules with an average diameter of 4 mm (range, 2–7 mm). Another five nodules were false-positive on PET/CT compared with seven nodules on dedicated CT. On the other hand, four of these seven false-positive nodules on dedicated CT (size: 4, 6, 8, and 11 mm) were correctly character-
TABLE 5: Osteosarcoma Follow-Up: Overall Diagnostic Performance No. of Lesions
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
91
89.3
72.2
83.1
81.3
82.4
MRI
9
80.0
100
100
80.0
88.9
Bone scanning
10
66.7
25.0
57.1
33.3
50.0
Examination CT
Note—PPV = positive predictive value, NPV = negative predictive value.
ized as true-negative by PET/CT because of their lack of FDG uptake. Bone lesions—A total of 16 metastatic and five benign skeletal lesions were detected; all 16 metastatic lesions were detected by PET/CT. Overall, PET/CT was superior to conventional imaging for characterizing skeletal lesions (accuracy = 95.2% vs 66.7%). Bone scanning had accuracy of only 50% in assessing a total of 10 findings, with three false-positive lesions (humeral head, ilium, and rib) and two false-negative lesions (distal humerus and femur). MRI had an accuracy of 85.7% in evaluating a total of seven bony sites. Lymph nodes and other sites of disease— A total of 16 abnormal lymph nodes were identified. Ten of these 16 nodes were metastatic, and all were correctly identified by both PET/CT and CT. False-positive FDG uptake was observed in one patient in inguinal and external iliac nodes. Abnormalities were identified at eight other sites by one of the imaging modalities, located in liver, skeletal muscles, breast, or pleura. PET/CT had a sensitivity of 100% for lesions in visceral organs, but the specificity was lower than it was for conventional imaging. The two false-positive findings recorded on PET/CT were located in the chest wall. In one patient, CT failed to identify a solitary liver metastasis that was detected by PET/CT. There were not enough MRI findings for meaningful analysis.
Fig. 2—Example of false-negative 18F-FDG PET in lung in 6-year-old boy with primary Ewing sarcoma of scapula. Staging low-dose CT (left) and fused PET/CT (right) axial images show subcentimeter lung nodule (arrow) suspicious for metastasis despite lack of FDG-avidity. There were several other nodules on other slices. Following resection of primary tumor and chemotherapy, patient underwent whole-lung radiotherapy for definitive treatment of suspected lung metastases.
Patient-Based Analysis and Effect on Management Among the 44 patients with Ewing sarcoma, PET/CT detected more metastatic lesions in nine patients and more accurately classified lesions in another 12 patients in at least one of the serial studies performed in these 21 patients. Similarly, among the 20 patients with osteosarcoma, PET/CT detected more metastatic lesions in six patients and more accurately classified lesions in another three patients in at least one of the serial studies obtained in these nine patients. Management changes that could be traced back to these imaging findings occurred in at least nine of the 64 patients, including initiation, direction, or avoidance of biopsies and initiation of radiotherapy or chemotherapy. It is difficult to determine the true effect of imaging findings on patient management in a retrospective analysis because studies were ordered by a variety of physicians, exact indications were not always well documented, and management decisions were influenced by a variety of factors (of which imaging was only one). Discussion The presence of metastatic disease is the single greatest adverse prognostic factor in pediatric patients with osteosarcoma and Ewing sarcoma [2, 5]. Accurate detection of metastatic lesions is thus important for accurate staging and selection of appropriate therapy [17]. Prior studies investigating the role of FDG PET/CT in patients with bone sarcoma were limited by the small size and heterogeneity (e.g., age, tumor type and tumor location) of the patient populations [18–25]. Importantly, the behavior and outcome of pediatric sarcomas often differs from that in adults [26]. For this reason, we chose to investigate the diagnostic performance of FDG PET/CT versus conventional imaging exclusively in pediatric patients with bone sarcomas. Our analysis suggests that FDG PET/CT has the highest value for evaluation of skeletal lesions; the accuracy in detecting malignant bone lesions was on par with that of
AJR:204, January 2015 157
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al.
A
B
C
D
Fig. 3—Example of true-positive FDG PET in lung: 20-year-old female with Ewing sarcoma. A–D, Surveillance images were obtained 6 years after initial treatment. Maximum-intensity-projection (MIP) (A), axial PET (B), low-dose CT (C), and fused PET/CT (D) images show hypermetabolic lung nodule in right upper lobe (arrow, A). MIP image also shows additional 18F-FDG-avid lesions in both lungs, multiple FDG-avid lymph nodes in left axilla, and incidental benign brown fat uptake in lower neck.
MRI. Both PET/CT and MRI were more accurate than were CT and bone scanning in the follow-up of both osteosarcoma and Ewing sarcoma, perhaps indicating that bone scanning could be eliminated in the followup of Ewing sarcoma and performed less frequently in osteosarcoma. One exception may be the small group of patients with predominantly osteosclerotic (rather than osteolytic) Ewing sarcoma because blastic osseous metastases generally show very low FDG uptake but can be detected on bone scanning [26]. FDG PET/CT is occasionally helpful in characterizing pulmonary metastases in pediatric sarcoma patients [25, 27], but in general, the sensitivity of PET/CT is lower than that of a dedicated chest CT. This is not surprising because the CT portion of the PET/CT
158
is usually performed during shallow breathing and at low resolution, limiting the detection of small lung nodules, in particular at the lung bases. In addition, partial volume effect secondary to respiratory motion during the PET diminishes the recorded activity concentration (SUV), in particular in small lung lesions. In our study, the size of the false-negative lung nodules was about 4 mm at staging and about 7.5 mm during follow-up. In particular, the three pulmonary metastases observed in our Ewing sarcoma staging group had very low FDG uptake (SUV, 0.4, 0.5, and 0.5) and measured between 2 and 6 mm. On the other hand, PET/CT may occasionally be helpful in characterizing doubtful lung nodules in sarcoma patients and may have a major role in assessing response to therapy [28, 29].
PET/CT and conventional imaging showed similar diagnostic performance for the detection of lymph node metastases. The relatively high proportion of false-positive findings on PET/CT in the follow-up of Ewing sarcoma is probably due to reactive changes in such nodes. For evaluation of the viscera and soft tissues, the overall accuracy of CT and MRI appears to be higher than that of PET/CT. However, the ability of PET/CT to detect lesions not seen by other modalities supports its use in patients at high risk. Combining the overall advantages of structural and functional imaging may be appropriate for performing a companion CT study of the PET/CT with full diagnostic dose parameters and IV contrast administration at least in patients at high risk, whereas low-dose CT
AJR:204, January 2015
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
PET/CT in Pediatric Bone Sarcoma without IV contrast administration may suffice in patients at low risk. On a per-scan basis, PET/CT overall identified a greater number of true-positive findings and a smaller number of false-positive findings compared with conventional imaging. This information may potentially translate into more accurate staging and better estimation of patient prognosis and may also be useful for tailoring treatment strategies. In fact, survival in bone sarcoma patients correlates with number and location of secondary lesions [9, 30], and it has been shown that patients with lung-only metastases have a longer survival compared with patients with bone involvement or a combination of pulmonary and osseous lesions [9]. It is difficult to compare our findings with previous studies because of the large heterogeneity in previous patient groups and the inconsistencies in the way data were reported. Our results are partially in keeping with those of Tateishi et al. [19] who showed a significant benefit of PET/CT over conventional imaging (chest radiography, diagnostic CT of the chest and abdomen, and locoregional MRI) in detecting distant metastases in a group of 50 pediatric patients with bone and soft-tissue sarcomas (including 20 with Ewing sarcoma and 18 with osteosarcoma) at initial staging. With regard to distant metastases analysis, conventional imaging understaged 15 patients (accuracy, 70%), whereas seven patients were understaged on PET/CT (accuracy, 86%). However, accuracy values for specific types of sarcoma were not reported in that article. In 2009, Piperkova et al. [31] assessed the diagnostic performance of FDG PET/CT in 93 patients with bone and soft-tissue sarcomas (15, osteosarcoma; two, Ewing sarcoma), comparing the results of combined PET/CT versus PET-only and CT. The combined PET/CT showed the best overall performance both for staging and restaging, with sensitivity and specificity of 100% (staging) and 100% and 96% (restaging), respectively. However, that study did not differentiate the results by histologic types of sarcomas and anatomic sites of metastases [31]. In a more recent study, London et al. [20] compared FDG PET/CT and conventional imaging in the detection of malignant lesions in 41 children with bone sarcomas. Similar to our findings, these authors documented high sensitivity, specificity, and accuracy for PET/CT (82%, 97%, and 96%, respectively). However, they did not provide any data regarding the diagnostic perfor-
mance of each individual conventional imaging modality and did not present data for different time points in the clinical management of these patients or for specific tumor histologic subtypes. Moreover, the statistical analysis placed particular emphasis on lung lesions and combined all lesions in other anatomic sites into a single dataset. The total number of follow-up PET/CT studies analyzed was considerably smaller than in the current study (86 vs 421). Our study has some limitations including the lack of available PET/CT for the initial staging of osteosarcoma and the relatively small proportion (67/412) of findings with histopathologic correlation. There may have been selection bias because not all imaging studies were performed at all time points. Nevertheless, our data provide a good foundation for planning future prospective studies in patients with pediatric bone sarcomas. It should also be noted that low FDG uptake in lung nodules does not necessarily preclude metastatic disease. We therefore used clinical and imaging follow-up for verification of lung nodules. Nevertheless, some uncertainty may remain because it is possible that lung nodules remain stable in size as long as patients are undergoing therapy. The diagnostic performance of imaging tests in pediatric patients needs to be weighed carefully against the radiation burden from such tests [32, 33]. The effective dose provided by using a standard PET/CT protocol ranges from 7.3 to 9.3 mSv (tube voltage, 120 kVp; tube current, 80 mAs; and abdominal pitch, 1.5:1). Potential dose savings could be realized by combining the dedicated CT and PET in one imaging study (thus eliminating the low-dose CT of the PET) or by combining PET/CT with a separate deep-inspiration breath-hold CT of the chest only (to address the issue of small lung nodules), by eliminating imaging tests with a low overall diagnostic yield (for instance, the low yield of bone scanning in our study), or in the future by using hybrid PET/MRI for staging and followup of pediatric sarcoma patients [34]. Regardless of these limitations, the results of this study have informed our patient management: Patients with newly diagnosed Ewing sarcoma now only undergo PET/CT at diagnosis and serially on posttherapy. Lesions identified on PET/CT are then further evaluated with MRI, dedicated CT, or other appropriate imaging. Biopsy or repeat imaging of equivocal findings is considered on a case-by-case basis with the expectation that
metastases will respond similarly to the primary site with effective systemic treatment or will progress with ineffective or no treatment. Given the different pattern of metastatic spread, the utility of PET/CT as a staging tool in osteosarcoma remains less clearly defined. Moreover, we see a need to evaluate the utility of this imaging modality as an early predictor of response to neoadjuvant chemotherapy. Conclusion Our data suggest that FDG PET/CT has good overall accuracy in the staging and follow-up of pediatric bone sarcomas. In concordance with the known patterns of metastatic recurrence (primarily in the lung in osteosarcoma and in lung and bone in Ewing sarcoma) the greater benefit was seen in Ewing sarcoma patients. Future prospective studies are needed to define the optimal imaging algorithms for initial staging and follow-up in these patients, including the number of studies and best time points for imaging during follow-up. Acknowledgments We thank Heather Magnan and Alexander Chou, both of Memorial Sloan Kettering Cancer Center, for clinical input and helpful discussion. References 1. Kaste SC. Imaging pediatric bone sarcomas. Radiol Clin North Am 2011; 49:749–765 2. Bakhshi S, Radhakrishnan V. Prognostic markers in osteosarcoma. Expert Rev Anticancer Ther 2010; 10:271–287 3. Treglia G, Salsano M, Stefanelli A, Mattoli MV, Giordano A, Bonomo L. Diagnostic accuracy of 18F-FDG-PET and PET/CT in patients with Ewing sarcoma family tumours: a systematic review and a meta-analysis. Skeletal Radiol 2012; 41:249–256 4. Wyers MR. Evaluation of pediatric bone lesions. Pediatr Radiol 2010; 40:468–473 5. Esiashvili N, Goodman M, Marcus RB Jr. Changes in incidence and survival of Ewing sarcoma patients over the past 3 decades: surveillance, epidemiology, and end results data. J Pediatr Hematol Oncol 2008; 30:425–430 6. Marina N, Gebhardt M, Teot L, Gorlick R. Biology and therapeutic advances for pediatric osteosarcoma. Oncologist 2004; 9:422–441 7. Quartuccio N, Cistaro A. Primary bone tumors. In: Cistara A, ed. Atlas of PET/CT in pediatric patients. Heidelberg, Germany: Springer, 2014 8. Quartuccio N, Treglia G, Salsano M, et al. The role of fluorine-18-fluorodeoxyglucose positron emission tomography in staging and restaging of patients
AJR:204, January 2015 159
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al. with osteosarcoma. Radiol Oncol 2013; 47:97–102 9. Cotterill SJ, Ahrens S, Paulussen M, et al. Prognostic factors in Ewing’s tumor of bone: analysis of 975 patients from the European Intergroup Cooperative Ewing’s Sarcoma Study Group. J Clin Oncol 2000; 18:3108–3114 10. Bernstein M, Kovar H, Paulussen M, et al. Ewing’s sarcoma family of tumors: current management. Oncologist 2006; 11:503–519 11. Quak E, van de Luijtgaarden AC, de Geus-Oei LF, van der Graaf WT, Oyen WJ. Clinical applications of positron emission tomography in sarcoma management. Expert Rev Anticancer Ther 2011; 11:195–204 12. Meyer JS, Nadel HR, Marina N, et al. Imaging guidelines for children with Ewing sarcoma and osteosarcoma: a report from the Children’s Oncology Group Bone Tumor Committee. Pediatr Blood Cancer 2008; 51:163–170 13. Weiser DA, Kaste SC, Siegel MJ, Adamson PC. Imaging in childhood cancer: A Society for Pediatric Radiology and Children’s Oncology group joint task force report. Pediatr Blood Cancer 2013; 60:1253–1260 14. Zasadny KR, Wahl RL. Standardized uptake values of normal tissues at PET with 2-[fluorine-18]fluoro-2-deoxy-D-glucose: variations with body weight and a method for correction. Radiology 1993; 189:847–850 15. Sugawara Y, Zasadny KR, Neuhoff AW, Wahl RL. Reevaluation of the standardized uptake value for FDG: variations with body weight and methods for correction. Radiology 1999; 213:521–525 16. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to PERCIST: evolving considerations for PET response criteria in solid tumors. J Nucl Med 2009; 50(suppl 1):122S–150S 17. ESMO / European Sarcoma Network Working
160
Group. Bone sarcomas: ESMO Clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol 2012; 23(suppl 7:vii100–vii109 18. Arush MW, Israel O, Postovsky S, et al. Positron emission tomography/computed tomography with 18fluoro-deoxyglucose in the detection of local recurrence and distant metastases of pediatric sarcoma. Pediatr Blood Cancer 2007; 49:901–905 19. Tateishi U, Hosono A, Makimoto A, et al. Accuracy of 18F fluorodeoxyglucose positron emission tomography/computed tomography in staging of pediatric sarcomas. J Pediatr Hematol Oncol 2007; 29:608–612 20. London K, Stege C, Cross S, et al. 18F-FDG PET/CT compared to conventional imaging modalities in pediatric primary bone tumors. Pediatr Radiol 2012; 42:418–430 21. Walter F, Czernin J, Hall T, et al. Is there a need for dedicated bone imaging in addition to 18F-FDG PET/CT imaging in pediatric sarcoma patients? J Pediatr Hematol Oncol 2012; 34:131–136 22. Iagaru A, Chawla S, Menendez L, Conti PS. 18FFDG PET and PET/CT for detection of pulmonary metastases from musculoskeletal sarcomas. Nucl Med Commun 2006; 27:795–802 23. Kleis M, Daldrup-Link H, Matthay K, et al. Diagnostic value of PET/CT for the staging and restaging of pediatric tumors. Eur J Nucl Med Mol Imaging 2009; 36:23–36 24. Fuglø HM, Jørgensen SM, Loft A, Hovgaard D, Petersen MM. The diagnostic and prognostic value of 18F-FDG PET/CT in the initial assessment of high-grade bone and soft tissue sarcoma: a retrospective study of 89 patients. Eur J Nucl Med Mol Imaging 2012; 39:1416–1424 25. Cistaro A, Lopci E, Gastaldo L, Fania P, Brach Del Prever A, Fagioli F. The role of 18F-FDG
PET/CT in the metabolic characterization of lung nodules in pediatric patients with bone sarcoma. Pediatr Blood Cancer 2012; 59:1206–1210 26. Maki RG. Pediatric sarcomas occurring in adults. J Surg Oncol 2008; 97:360–368 27. Ulaner GA, Magnan H, Healey JH, Weber WA, Meyers PA. Is methylene diphosphonate bone scan necessary for initial staging of Ewing sarcoma if FDG PET/CT is performed? AJR 2014; 202:859–867 28. Benz MR, Czernin J, Tap WD, let al. FDGPET/CT imaging predicts histopathologic treatment responses after neoadjuvant therapy in adult primary bone sarcomas. Sarcoma 2010; 2010:143540 29. Byun BH, Kong CB, Lim I, et al.. Early response monitoring to neoadjuvant chemotherapy in osteosarcoma using sequential 18F-FDG PET/CT and MRI. Eur J Nucl Med Mol Imaging 2014; 41:1553–1562 30. Kager L, Zoubek A, Pötschger U, et al; Cooperative German-Austrian-Swiss Osteosarcoma Study Group. Primary metastatic osteosarcoma: presentation and outcome of patients treated on neoadjuvant Cooperative Osteosarcoma Study Group protocols. J Clin Oncol 2003; 21:2011–2018 31. Piperkova E, Mikhaeil M, Mousavi A, et al. Impact of PET and CT in PET/CT studies for staging and evaluating treatment response in bone and soft tissue sarcomas. Clin Nucl Med 2009; 34:146–150 32. Huda W, Vance A. Patient radiation doses from adult and pediatric CT. AJR 2007; 188:540–546 33. Fahey FH. Dosimetry of pediatric PET/CT. J Nucl Med 2009; 50:1483–1491 34. Buchbender C, Heusner TA, Lauenstein TC, Bockisch A, Antoch G. Oncologic PET/MRI. Part 2. Bone tumors, soft-tissue tumors, melanoma, and lymphoma. J Nucl Med 2012; 53:1244–1252
AJR:204, January 2015
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Nuclear Medicine and Molecular Imaging Original Research
Pediatric Bone Sarcoma: Diagnostic Performance of 18F-FDG PET/CT Versus Conventional Imaging for Initial Staging and Follow-Up Natale Quartuccio1 Josef Fox 2 Deborah Kuk 3 Leonard H. Wexler4 Sergio Baldari1 Angelina Cistaro5,6,7 Heiko Schöder 2 Quartuccio N, Fox J, Kuk D, et al. Keywords: 18F-FDG PET/CT, Ewing sarcoma, osteosarcoma, pediatric bone sarcoma, prognosis DOI:10.2214/AJR.14.12932 Received March 28, 2014; accepted after revision July 3, 2014. N. Quartuccio was a recipient of the 2011 Bradley-Alavi Student Fellowship Award. N. Quartuccio and A. Cistaro contributed equally to this work. 1
Nuclear Medicine Unit, Department of Biomedical Sciences and of Morphologic and Functional Images, University of Messina, Messina, Italy.
2
Molecular Imaging and Therapy Service, Department of Radiology, Memorial Sloan Kettering Cancer Center, 1275 York Ave, New York, NY 10061. Address correspondence to H. Schöder ([email protected]). 3 Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, NY. 4 Department of Pediatrics, Memorial Sloan Kettering Cancer Center, New York, NY. 5
PET Centre IRMET S.p.A., Euromedic, Turin, Italy.
6
Coordinator of PET Pediatric AIMN InterGroup, Turin, Italy.
7
Institute of Cognitive Sciences and Technologies, National Research Council, Rome, Italy.
AJR 2015; 204:153–160 0361–803X/15/2041–153 © American Roentgen Ray Society
OBJECTIVE. The purpose of this study was to compare the diagnostic performance of PET/CT and conventional imaging for staging and follow-up of pediatric osteosarcoma and skeletal Ewing sarcoma. MATERIALS AND METHODS. We calculated sensitivity, specificity, and accuracy of PET/CT and conventional imaging (CT, MRI, bone scanning) for sites of disease and number of lesions. Diagnostic benefit, defined as better characterization of lesions, was evaluated on a per-scan basis, comparing PET/CT and conventional imaging. RESULTS. A total of 412 lesions were characterized by imaging in 64 patients (20, osteosarcoma; 44, Ewing sarcoma). For osteosarcoma patients PET/CT was available only at follow-up, where it proved more accurate than conventional imaging for the detection of bone lesions (accuracy, 95% vs 67% for CT and 86% for MRI) and complementary to CT in evaluating lung nodules (sensitivity, 84% vs 94%; specificity, 79% vs 71%) with diagnostic benefit in 18% of examinations. In patients with Ewing sarcoma, PET/CT tended to perform better during follow-up than at initial staging (accuracy, 85% vs 69%). For lung findings, PET/CT was more specific than CT but was less sensitive. The diagnostic benefit of PET/CT was greater at staging (28%) than during followup (9%). On a per-patient basis, PET/CT provided diagnostic benefit in 21 of 44 patients with Ewing sarcoma and nine of 20 patients with osteosarcoma at least once during clinical management. CONCLUSION. FDG PET/CT provides diagnostic benefit in Ewing sarcoma and osteosarcoma, with the exception of small lung nodules. Prospective studies are needed to define the best imaging algorithm and combination of tests in the staging and follow-up of patients with pediatric bone sarcoma. 18F-FDG
T
he most common primary malignant bone tumors in children are osteosarcoma and Ewing sarcoma [1]. Osteosarcoma derives from primitive bone-forming mesenchymal stem cells and occurs mainly in the metaphyseal portion of the long bones [2]. Ewing sarcoma is equally divided between the axial and the appendicular skeleton [3, 4]. Both tumors have a peak incidence in the second decade of life, and the presence of metastases at diagnosis is one of the strongest predictors of survival [2, 5]. Approximately 20–25% of patients with osteosarcoma and Ewing sarcoma present with radiographically detectable distant metastases [6, 7]; however, the proportion of patients with lung-only metastases is substantially greater in osteosarcoma than in Ewing sarcoma, in which bone or bone marrow metastases are seen alone or in combination with lung metastases in nearly one half of patients with distant metastases
[8, 9]. Lymph node and visceral metastases are uncommon in both tumor types at diagnosis but may be seen in the setting of advanced, recurrent disease [6, 10]. The 5-year survival for both malignancies is 65–70% with localized disease compared with only 20% with detectable metastatic disease [2, 6]. Accurate detection of metastases is important for risk stratification and may influence the clinical management of these patients. An important imaging modality for assessing bone sarcomas is 18F-FDG PET/CT [11]. Few studies have investigated the diagnostic performance of FDG PET/CT in children with osteosarcoma and Ewing sarcoma [8–11]. Presently, FDG PET/CT is not considered a standard examination in the diagnostic algorithm for these patients [12, 13]. Nevertheless, in our clinical experience FDG PET/CT accurately depicts metastatic lesions from bone sarcoma both at diagnosis and at the time of relapse. The purpose of
AJR:204, January 2015 153
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al. this study was to compare the performance of FDG PET/CT with conventional imaging for detecting osteosarcoma and Ewing sarcoma lesions in different anatomic regions. Additionally, we aimed to determine the potential impact of FDG PET/CT results on patient management. Materials and Methods Patients This retrospective study was approved by our institutional review board with a waiver of authorization. Patient consent was not required. A retrospective search of our institutional PET database from February 2002 to September 2012 was conducted using the following criteria: age more than 18 years at diagnosis; biopsy-proven diagnosis of osteosarcoma or skeletal Ewing sarcoma; availability of at least one FDG PET/CT study for tumor staging or follow-up at our institution; and availability of one or more conventional imaging modalities, including dedicated CT, MRI, or bone scanning performed within 1 month of the comparison PET/CT examination. For evaluation of pulmonary nodules, the availability of a dedicated chest CT study (with or without IV contrast material) was required as the comparator. Bone lesions were assessed using all imaging techniques. Lymph node metastases were evaluated by MRI and CT. Other sites (such as soft tissue, liver, and brain) were evaluated by the most relevant conventional imaging modality. Imaging datasets and reports of PET/CT and conventional imaging were reviewed. Histopathologic examination of lesions (if available) or clinical and imaging follow-up for at least 6 months were considered the standard of reference.
PET/CT Examination and Image Analysis PET/CT studies were performed on various systems (Discovery LS, Discovery ST, Discovery STE, or Discovery PET/CT 610, all GE Healthcare). Patients fasted for at least 6 hours before scanning. Plasma glucose tested before FDG injection was less than 160 mg/dL in all patients. PET/CT scans were acquired in 2D mode before 2008 and in 3D mode thereafter. The average uptake time was 72 minutes; injected activity ranged from 113 to 596 MBq, depending on the body surface area. Images were reconstructed using dedicated software. Scans were visually inspected for the presence of abnormal uptake distinguishable from background activity and suggestive of tumor metabolism. Maximum SUV values (SUVmax) normalized for body weight were then measured for each lesion [14, 15]. Reference background FDG activity was measured in the right hepatic lobe and
154
TABLE 1: Patient Characteristics Characteristics
All Patients
Ewing Sarcoma
Osteosarcoma
64
44
20
Median age at diagnosis (y)
13.2
13.3
13.1
Median age at PET/CT (y)
15.0
14.7
15.5 10
No. of patients
Female
34
24
Male
30
20
10
421
346
75
Total no. PET/CT No. of PET/CT at staging
18
18
0
No. of PET/CT at follow-up
403
328
75
Total no. of lesions found
412
311
101
Mean SUV liver
2.3
2.3
2.41
Mean SUV blood pool
1.6
1.6
1.7
Background regions
Primary lesions Mean SUVmax (g/L)
4.9 (1.9–10.8)
5.0 (1.9–10.8)
NA
Mean dimensions (mm)
49.9 (2.9–175)
52.7 (2.9–175)
NA
Mean SUVmax range (g/L)
2.3 (0.2–6.3)
0.8 (0.2–5.6)
2.9 (0.4–6.3)
Mean dimensions range (mm)
9 (1.7–59.7)
6.9 (1.7–48.8)
19 (2.0–59.7)
Lung lesions
Note—Data in parentheses are range. SUVmax = maximum SUV. NA = not applicable.
mediastinal blood pool [16]. Lesion size was measured on MRI or CT, depending on the location.
Statistical Analysis Patient and scanning characteristics are presented using the median, mean, and range for continuous variables and counts for categoric variables. FDG PET/CT studies were compared with CT, MRI and bone scanning studies in a lesionbased (pulmonary, bone, lymph node, or other) analysis at initial staging and follow-up and for each disease. According to the standard of reference, imaging findings were retrospectively classified as true-positive, true-negative, false-positive, or false-negative. Findings described in the reports as “equivocal,” “not conclusive,” or “suggestive of malignancy” were considered PETpositive and then classified as true-positive or false-positive lesions using the same standard of reference. Concordance between PET/CT and conventional imaging results was assessed. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated on a lesional and patient basis and categorized by time point (initial staging versus follow-up). Finally, we investigated whether PET/CT results provided a diagnostic benefit beyond that of conventional imaging. This was determined by comparing the number of true-positive and false-positive lesions recorded per scan; recording of more true-positive lesions or fewer
false-positive lesions on PET versus conventional imaging represented a diagnostic benefit. All statistical analyses were performed using R, version 3.0.2 (R Project). A p value of less than 0.05 was considered significant.
Results We retrieved 421 PET/CT examinations. A total of 412 lesions were compared on FDG PET/CT and conventional imaging in a total of 64 patients (Table 1). Pathology results were available for 67 of 412 findings (20, pulmonary; 47, nonpulmonary); a minimum 6 months of follow-up was used for confirmation of the remaining cases. This analysis includes 16 primary Ewing sarcomas; of the remaining 396 lesions, 271 were eventually classified as metastasis on the basis of the standard of reference. Ewing Sarcoma A total of 311 lesions (lung, 132; bone, 115; lymph node, 32; and other sites, 32) were characterized on imaging with an overall concordance of 79.2% (staging, 68.8%; follow-up, 80.9%) between FDG PET/CT and conventional imaging (Tables 2–4). Primary tumors (n = 16) showed variable FDG uptake, with SUVmax between 1.9 and 10.8 (Table 1). In the per-scan comparison, a diagnostic benefit was seen in 27.8% of initial
AJR:204, January 2015
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
staging PET/CT studies (Fig. 1), with fewer false-positive lesions compared with CT in two scan-sets and more true-positive lesions in three scan-sets. For follow-up scans, a diagnostic benefit was seen in 9.1% of cases (30/328 datasets), including fewer false-positive findings in 17 examination sets and more true-positive lesions in 13 examination sets. Lung lesions—At initial staging, all three malignant nodules were correctly classified on dedicated chest CT but not on PET/CT (lesion size, 6.1, 2.3, and 5.6 mm; SUV, 0.4, 0.5, and 0.5, respectively). At follow-up, 25 of the 28 malignant nodules were detected on chest CT (mean size, 6.9 mm; range, 1.7– 25.5 mm). The average SUV of these nodules was 1.0. Four of the 28 nodules were false-negative on PET/CT (size range, 1.7– 3.5 mm; all with SUVmax ≤ 0.2). In comparison with CT, PET/CT had a slightly lower sensitivity (85.7% vs 89.3%) but a slightly higher specificity (90.7% vs 84.9%). Representative cases are shown in Figures 2 and 3. Bone lesions—At initial staging, a total of 20 metastatic lesions, distant from the primary tumor, were identified. For lesions within the FOV of each imaging test, MRI showed sensitivity of 100% and the highest overall accuracy (88.9%). CT performed the worst, detecting only four of nine metastases. Whole-body PET/CT detected all 16 malignant lesions (sensitivity 100%) but at the expense of a low specificity of 25%. At follow-up, a total of 95 bone lesions (10 suspected local recurrences and 85 distant sites) were identified, of which 50 were malignant. In the overall assessment, PET/CT showed slightly higher sensitivity than did MRI, with similar specificity. Among the 10 suspected local recurrences (on the basis of clinical examination or findings in one of the imaging tests) only three were found to be true. These three lesions were detected by both PET and bone scanning, but the bone scanning study showed more false-positive sites. Two of the three lesions were also positive on MRI; the third lesion was not imaged with MRI. Among the eight local recurrence sites with available PET and MRI, imaging findings were concordant for six and discordant for two (false-positive that was true-negative or true-negative that was false-positive result). Lymph nodes—A total of 32 nodal lesions were identified either at initial staging (n = 4) or during follow-up (n = 28). At staging, all metastatic nodes were correctly identified by PET/CT or CT/MRI. At follow-up, eight FDGavid nodes proved to be false-positive (four in the mediastinum and one each in the subpectoral, axillary, cervical, and inguinal regions). MRI also showed two false-positive nodes.
TABLE 2: Ewing Sarcoma Overall Analysis: Lesion-Based Diagnostic PET/CT in Pediatric Bone Sarcoma Performance Combined for Staging and Follow-Up No. of Lesions
Sensitivity (%)
PET/CT
311
90.3
CT
231
71.1
MRI
95
92.1
77.2
Bone scanning
91
42.3
66.7
Examination
Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
77.8
69.9
93.3
82
84.5
69.2
85.6
80.1
72.9
93.6
83.2
62.9
46.4
52.8
Note—PPV = positive predictive value, NPV = negative predictive value.
TABLE 3: Ewing Sarcoma Staging: Diagnostic Performance by Disease Site Lesion
No. Sensitivity (%) Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
Lung PET/CT
18
0
53.3
0
72.7
44.4
CT
18
100
46.7
27.3
100
55.6
20
100
25
84.2
100
85
Bone PET/CT CT
9
44.4
Not a numbera
100
0
44.4
MRI
9
100
75
83.3
100
88.9
Bone scanning
20
62.5
100
100
40
70
4
100
66.7
50
100
75
CT
3
Not a numbera
100
Not a numbera
100
100
MRI
4
100
100
100
100
100
3
100
100
100
100
100
CT
2
Not a numbera
100
Not a numbera
100
100
MRI
1
100
Not a numbera
100
Not a numbera
100
Lymph node PET/CT
Other PET/CT
Note—PPV = positive predictive value, NPV = negative predictive value, NA = not assessable. aOccurs when the denominator is zero.
TABLE 4: Ewing Sarcoma Follow-Up: Diagnostic Performance by Disease Site Lesion
No. Sensitivity (%) Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
Lung PET/CT
114
85.7
90.7
75
95.1
89.5
CT
114
89.3
84.9
65.8
96.1
86.0 86.3
Bone PET/CT
95
96.0
75.6
81.4
94.4
CT
46
57.1
83.3
84.2
55.6
67.4
MRI
56
88.0
74.2
73.3
88.5
80.4
Bone scanning
71
33.3
62.9
48.0
47.8
47.9
Lymph node PET/CT
28
100
66.7
33.3
100
71.4
CT
22
75.0
100
100
94.7
95.6
MRI
7
Not a numbera
71.4
0
100
71.4
PET/CT
29
80.0
68.4
57.1
86.7
72.4
CT
17
75.0
100
100
92.9
94.1
MRI
18
100
83.3
75.0
100
88.9
Other
Note—PPV = positive predictive value, NPV = negative predictive value. aOccurs when the denominator is zero.
AJR:204, January 2015 155
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al.
A
B
Fig. 1—Example of diagnostic benefit of 18F-FDG PET at staging in 14-year-old girl with primary Ewing sarcoma of right femur. A, Axial low-dose CT (left) and fused PET/CT (right) images show extensive hypermetabolic mass involving right proximal femur and thigh. B, Coronal low-dose CT (left), fused PET/CT (center), and PET (right) images clearly depict extension of hypermetabolic tumor into right femoral vein (arrow). C, Maximum-intensity-projection (left) and fused axial PET/CT (right) images show additional FDGavid lesion within right ventricle of heart (arrows) suspicious for tumor thrombus in view of right femoral vein involvement. D, Axial T1 fat-saturated gadolinium-enhanced cardiac MR image obtained 1 day later confirmed mobile pedunculated mass (arrow) adjacent to free wall of right ventricle. At surgery, 3 × 2 × 1 cm mass was resected, yielding metastatic Ewing sarcoma. E, Intracardiac lesion was not detected prospectively on contrast-enhanced CT performed 1 week before PET/CT, although right ventricular filling defect is evident on retrospective review. In either case, FDGavidity on PET enabled diagnosis of tumor thrombus (as opposed to bland thrombus).
156
C
D
E AJR:204, January 2015
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
PET/CT in Pediatric Bone Sarcoma Other sites—There were a total of 32 abnormal findings in other body regions at initial staging (liver, n = 2; skeletal muscle, n = 1) and during follow-up (liver, n = 3; skeletal muscle, n = 21; and n = 1 each for brain, heart, scalp, nasopharynx, and pleura). At initial staging, there was no difference in diagnostic performance between PET/CT and conventional imaging; both showed an accuracy of 100%. During follow-up, MRI had the highest sensitivity (100%, n = 18) and CT had the highest specificity (100%, n = 17). On PET/CT, two lesions were false-negative, one in the paraspinous soft tissues and one in the left forearm. Osteosarcoma Only follow-up PET/CT examinations were available for osteosarcoma patients (Table 5). Concordance of imaging findings on PET/CT versus conventional imaging was noted in 82.4% of cases. Diagnostic benefit was provided by PET/CT in 17.4% of datasets, showing fewer false-positive lesions in 6.7% (5/75) and more true-positive lesions in 10.7% (8/75). Lung lesions—A total of 32 malignant and 24 benign nodules were detected. The mean size of the malignant nodules was 19 mm (range, 2–60 mm); the mean SUV was 2.9 (range, 0.4–6.3). Dedicated CT had a lower rate of false-negative findings than did PET/CT. The two nodules that were falsely reported as benign on CT measured 5 and 2 mm. PET/CT failed to correctly characterize five malignant lung nodules with an average diameter of 4 mm (range, 2–7 mm). Another five nodules were false-positive on PET/CT compared with seven nodules on dedicated CT. On the other hand, four of these seven false-positive nodules on dedicated CT (size: 4, 6, 8, and 11 mm) were correctly character-
TABLE 5: Osteosarcoma Follow-Up: Overall Diagnostic Performance No. of Lesions
Sensitivity (%)
Specificity (%)
PPV (%)
NPV (%)
Accuracy (%)
91
89.3
72.2
83.1
81.3
82.4
MRI
9
80.0
100
100
80.0
88.9
Bone scanning
10
66.7
25.0
57.1
33.3
50.0
Examination CT
Note—PPV = positive predictive value, NPV = negative predictive value.
ized as true-negative by PET/CT because of their lack of FDG uptake. Bone lesions—A total of 16 metastatic and five benign skeletal lesions were detected; all 16 metastatic lesions were detected by PET/CT. Overall, PET/CT was superior to conventional imaging for characterizing skeletal lesions (accuracy = 95.2% vs 66.7%). Bone scanning had accuracy of only 50% in assessing a total of 10 findings, with three false-positive lesions (humeral head, ilium, and rib) and two false-negative lesions (distal humerus and femur). MRI had an accuracy of 85.7% in evaluating a total of seven bony sites. Lymph nodes and other sites of disease— A total of 16 abnormal lymph nodes were identified. Ten of these 16 nodes were metastatic, and all were correctly identified by both PET/CT and CT. False-positive FDG uptake was observed in one patient in inguinal and external iliac nodes. Abnormalities were identified at eight other sites by one of the imaging modalities, located in liver, skeletal muscles, breast, or pleura. PET/CT had a sensitivity of 100% for lesions in visceral organs, but the specificity was lower than it was for conventional imaging. The two false-positive findings recorded on PET/CT were located in the chest wall. In one patient, CT failed to identify a solitary liver metastasis that was detected by PET/CT. There were not enough MRI findings for meaningful analysis.
Fig. 2—Example of false-negative 18F-FDG PET in lung in 6-year-old boy with primary Ewing sarcoma of scapula. Staging low-dose CT (left) and fused PET/CT (right) axial images show subcentimeter lung nodule (arrow) suspicious for metastasis despite lack of FDG-avidity. There were several other nodules on other slices. Following resection of primary tumor and chemotherapy, patient underwent whole-lung radiotherapy for definitive treatment of suspected lung metastases.
Patient-Based Analysis and Effect on Management Among the 44 patients with Ewing sarcoma, PET/CT detected more metastatic lesions in nine patients and more accurately classified lesions in another 12 patients in at least one of the serial studies performed in these 21 patients. Similarly, among the 20 patients with osteosarcoma, PET/CT detected more metastatic lesions in six patients and more accurately classified lesions in another three patients in at least one of the serial studies obtained in these nine patients. Management changes that could be traced back to these imaging findings occurred in at least nine of the 64 patients, including initiation, direction, or avoidance of biopsies and initiation of radiotherapy or chemotherapy. It is difficult to determine the true effect of imaging findings on patient management in a retrospective analysis because studies were ordered by a variety of physicians, exact indications were not always well documented, and management decisions were influenced by a variety of factors (of which imaging was only one). Discussion The presence of metastatic disease is the single greatest adverse prognostic factor in pediatric patients with osteosarcoma and Ewing sarcoma [2, 5]. Accurate detection of metastatic lesions is thus important for accurate staging and selection of appropriate therapy [17]. Prior studies investigating the role of FDG PET/CT in patients with bone sarcoma were limited by the small size and heterogeneity (e.g., age, tumor type and tumor location) of the patient populations [18–25]. Importantly, the behavior and outcome of pediatric sarcomas often differs from that in adults [26]. For this reason, we chose to investigate the diagnostic performance of FDG PET/CT versus conventional imaging exclusively in pediatric patients with bone sarcomas. Our analysis suggests that FDG PET/CT has the highest value for evaluation of skeletal lesions; the accuracy in detecting malignant bone lesions was on par with that of
AJR:204, January 2015 157
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al.
A
B
C
D
Fig. 3—Example of true-positive FDG PET in lung: 20-year-old female with Ewing sarcoma. A–D, Surveillance images were obtained 6 years after initial treatment. Maximum-intensity-projection (MIP) (A), axial PET (B), low-dose CT (C), and fused PET/CT (D) images show hypermetabolic lung nodule in right upper lobe (arrow, A). MIP image also shows additional 18F-FDG-avid lesions in both lungs, multiple FDG-avid lymph nodes in left axilla, and incidental benign brown fat uptake in lower neck.
MRI. Both PET/CT and MRI were more accurate than were CT and bone scanning in the follow-up of both osteosarcoma and Ewing sarcoma, perhaps indicating that bone scanning could be eliminated in the followup of Ewing sarcoma and performed less frequently in osteosarcoma. One exception may be the small group of patients with predominantly osteosclerotic (rather than osteolytic) Ewing sarcoma because blastic osseous metastases generally show very low FDG uptake but can be detected on bone scanning [26]. FDG PET/CT is occasionally helpful in characterizing pulmonary metastases in pediatric sarcoma patients [25, 27], but in general, the sensitivity of PET/CT is lower than that of a dedicated chest CT. This is not surprising because the CT portion of the PET/CT
158
is usually performed during shallow breathing and at low resolution, limiting the detection of small lung nodules, in particular at the lung bases. In addition, partial volume effect secondary to respiratory motion during the PET diminishes the recorded activity concentration (SUV), in particular in small lung lesions. In our study, the size of the false-negative lung nodules was about 4 mm at staging and about 7.5 mm during follow-up. In particular, the three pulmonary metastases observed in our Ewing sarcoma staging group had very low FDG uptake (SUV, 0.4, 0.5, and 0.5) and measured between 2 and 6 mm. On the other hand, PET/CT may occasionally be helpful in characterizing doubtful lung nodules in sarcoma patients and may have a major role in assessing response to therapy [28, 29].
PET/CT and conventional imaging showed similar diagnostic performance for the detection of lymph node metastases. The relatively high proportion of false-positive findings on PET/CT in the follow-up of Ewing sarcoma is probably due to reactive changes in such nodes. For evaluation of the viscera and soft tissues, the overall accuracy of CT and MRI appears to be higher than that of PET/CT. However, the ability of PET/CT to detect lesions not seen by other modalities supports its use in patients at high risk. Combining the overall advantages of structural and functional imaging may be appropriate for performing a companion CT study of the PET/CT with full diagnostic dose parameters and IV contrast administration at least in patients at high risk, whereas low-dose CT
AJR:204, January 2015
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
PET/CT in Pediatric Bone Sarcoma without IV contrast administration may suffice in patients at low risk. On a per-scan basis, PET/CT overall identified a greater number of true-positive findings and a smaller number of false-positive findings compared with conventional imaging. This information may potentially translate into more accurate staging and better estimation of patient prognosis and may also be useful for tailoring treatment strategies. In fact, survival in bone sarcoma patients correlates with number and location of secondary lesions [9, 30], and it has been shown that patients with lung-only metastases have a longer survival compared with patients with bone involvement or a combination of pulmonary and osseous lesions [9]. It is difficult to compare our findings with previous studies because of the large heterogeneity in previous patient groups and the inconsistencies in the way data were reported. Our results are partially in keeping with those of Tateishi et al. [19] who showed a significant benefit of PET/CT over conventional imaging (chest radiography, diagnostic CT of the chest and abdomen, and locoregional MRI) in detecting distant metastases in a group of 50 pediatric patients with bone and soft-tissue sarcomas (including 20 with Ewing sarcoma and 18 with osteosarcoma) at initial staging. With regard to distant metastases analysis, conventional imaging understaged 15 patients (accuracy, 70%), whereas seven patients were understaged on PET/CT (accuracy, 86%). However, accuracy values for specific types of sarcoma were not reported in that article. In 2009, Piperkova et al. [31] assessed the diagnostic performance of FDG PET/CT in 93 patients with bone and soft-tissue sarcomas (15, osteosarcoma; two, Ewing sarcoma), comparing the results of combined PET/CT versus PET-only and CT. The combined PET/CT showed the best overall performance both for staging and restaging, with sensitivity and specificity of 100% (staging) and 100% and 96% (restaging), respectively. However, that study did not differentiate the results by histologic types of sarcomas and anatomic sites of metastases [31]. In a more recent study, London et al. [20] compared FDG PET/CT and conventional imaging in the detection of malignant lesions in 41 children with bone sarcomas. Similar to our findings, these authors documented high sensitivity, specificity, and accuracy for PET/CT (82%, 97%, and 96%, respectively). However, they did not provide any data regarding the diagnostic perfor-
mance of each individual conventional imaging modality and did not present data for different time points in the clinical management of these patients or for specific tumor histologic subtypes. Moreover, the statistical analysis placed particular emphasis on lung lesions and combined all lesions in other anatomic sites into a single dataset. The total number of follow-up PET/CT studies analyzed was considerably smaller than in the current study (86 vs 421). Our study has some limitations including the lack of available PET/CT for the initial staging of osteosarcoma and the relatively small proportion (67/412) of findings with histopathologic correlation. There may have been selection bias because not all imaging studies were performed at all time points. Nevertheless, our data provide a good foundation for planning future prospective studies in patients with pediatric bone sarcomas. It should also be noted that low FDG uptake in lung nodules does not necessarily preclude metastatic disease. We therefore used clinical and imaging follow-up for verification of lung nodules. Nevertheless, some uncertainty may remain because it is possible that lung nodules remain stable in size as long as patients are undergoing therapy. The diagnostic performance of imaging tests in pediatric patients needs to be weighed carefully against the radiation burden from such tests [32, 33]. The effective dose provided by using a standard PET/CT protocol ranges from 7.3 to 9.3 mSv (tube voltage, 120 kVp; tube current, 80 mAs; and abdominal pitch, 1.5:1). Potential dose savings could be realized by combining the dedicated CT and PET in one imaging study (thus eliminating the low-dose CT of the PET) or by combining PET/CT with a separate deep-inspiration breath-hold CT of the chest only (to address the issue of small lung nodules), by eliminating imaging tests with a low overall diagnostic yield (for instance, the low yield of bone scanning in our study), or in the future by using hybrid PET/MRI for staging and followup of pediatric sarcoma patients [34]. Regardless of these limitations, the results of this study have informed our patient management: Patients with newly diagnosed Ewing sarcoma now only undergo PET/CT at diagnosis and serially on posttherapy. Lesions identified on PET/CT are then further evaluated with MRI, dedicated CT, or other appropriate imaging. Biopsy or repeat imaging of equivocal findings is considered on a case-by-case basis with the expectation that
metastases will respond similarly to the primary site with effective systemic treatment or will progress with ineffective or no treatment. Given the different pattern of metastatic spread, the utility of PET/CT as a staging tool in osteosarcoma remains less clearly defined. Moreover, we see a need to evaluate the utility of this imaging modality as an early predictor of response to neoadjuvant chemotherapy. Conclusion Our data suggest that FDG PET/CT has good overall accuracy in the staging and follow-up of pediatric bone sarcomas. In concordance with the known patterns of metastatic recurrence (primarily in the lung in osteosarcoma and in lung and bone in Ewing sarcoma) the greater benefit was seen in Ewing sarcoma patients. Future prospective studies are needed to define the optimal imaging algorithms for initial staging and follow-up in these patients, including the number of studies and best time points for imaging during follow-up. Acknowledgments We thank Heather Magnan and Alexander Chou, both of Memorial Sloan Kettering Cancer Center, for clinical input and helpful discussion. References 1. Kaste SC. Imaging pediatric bone sarcomas. Radiol Clin North Am 2011; 49:749–765 2. Bakhshi S, Radhakrishnan V. Prognostic markers in osteosarcoma. Expert Rev Anticancer Ther 2010; 10:271–287 3. Treglia G, Salsano M, Stefanelli A, Mattoli MV, Giordano A, Bonomo L. Diagnostic accuracy of 18F-FDG-PET and PET/CT in patients with Ewing sarcoma family tumours: a systematic review and a meta-analysis. Skeletal Radiol 2012; 41:249–256 4. Wyers MR. Evaluation of pediatric bone lesions. Pediatr Radiol 2010; 40:468–473 5. Esiashvili N, Goodman M, Marcus RB Jr. Changes in incidence and survival of Ewing sarcoma patients over the past 3 decades: surveillance, epidemiology, and end results data. J Pediatr Hematol Oncol 2008; 30:425–430 6. Marina N, Gebhardt M, Teot L, Gorlick R. Biology and therapeutic advances for pediatric osteosarcoma. Oncologist 2004; 9:422–441 7. Quartuccio N, Cistaro A. Primary bone tumors. In: Cistara A, ed. Atlas of PET/CT in pediatric patients. Heidelberg, Germany: Springer, 2014 8. Quartuccio N, Treglia G, Salsano M, et al. The role of fluorine-18-fluorodeoxyglucose positron emission tomography in staging and restaging of patients
AJR:204, January 2015 159
Downloaded from www.ajronline.org by SUNY Downstate Medical Center on 03/23/15 from IP address 138.5.159.110. Copyright ARRS. For personal use only; all rights reserved
Quartuccio et al. with osteosarcoma. Radiol Oncol 2013; 47:97–102 9. Cotterill SJ, Ahrens S, Paulussen M, et al. Prognostic factors in Ewing’s tumor of bone: analysis of 975 patients from the European Intergroup Cooperative Ewing’s Sarcoma Study Group. J Clin Oncol 2000; 18:3108–3114 10. Bernstein M, Kovar H, Paulussen M, et al. Ewing’s sarcoma family of tumors: current management. Oncologist 2006; 11:503–519 11. Quak E, van de Luijtgaarden AC, de Geus-Oei LF, van der Graaf WT, Oyen WJ. Clinical applications of positron emission tomography in sarcoma management. Expert Rev Anticancer Ther 2011; 11:195–204 12. Meyer JS, Nadel HR, Marina N, et al. Imaging guidelines for children with Ewing sarcoma and osteosarcoma: a report from the Children’s Oncology Group Bone Tumor Committee. Pediatr Blood Cancer 2008; 51:163–170 13. Weiser DA, Kaste SC, Siegel MJ, Adamson PC. Imaging in childhood cancer: A Society for Pediatric Radiology and Children’s Oncology group joint task force report. Pediatr Blood Cancer 2013; 60:1253–1260 14. Zasadny KR, Wahl RL. Standardized uptake values of normal tissues at PET with 2-[fluorine-18]fluoro-2-deoxy-D-glucose: variations with body weight and a method for correction. Radiology 1993; 189:847–850 15. Sugawara Y, Zasadny KR, Neuhoff AW, Wahl RL. Reevaluation of the standardized uptake value for FDG: variations with body weight and methods for correction. Radiology 1999; 213:521–525 16. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to PERCIST: evolving considerations for PET response criteria in solid tumors. J Nucl Med 2009; 50(suppl 1):122S–150S 17. ESMO / European Sarcoma Network Working
160
Group. Bone sarcomas: ESMO Clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol 2012; 23(suppl 7:vii100–vii109 18. Arush MW, Israel O, Postovsky S, et al. Positron emission tomography/computed tomography with 18fluoro-deoxyglucose in the detection of local recurrence and distant metastases of pediatric sarcoma. Pediatr Blood Cancer 2007; 49:901–905 19. Tateishi U, Hosono A, Makimoto A, et al. Accuracy of 18F fluorodeoxyglucose positron emission tomography/computed tomography in staging of pediatric sarcomas. J Pediatr Hematol Oncol 2007; 29:608–612 20. London K, Stege C, Cross S, et al. 18F-FDG PET/CT compared to conventional imaging modalities in pediatric primary bone tumors. Pediatr Radiol 2012; 42:418–430 21. Walter F, Czernin J, Hall T, et al. Is there a need for dedicated bone imaging in addition to 18F-FDG PET/CT imaging in pediatric sarcoma patients? J Pediatr Hematol Oncol 2012; 34:131–136 22. Iagaru A, Chawla S, Menendez L, Conti PS. 18FFDG PET and PET/CT for detection of pulmonary metastases from musculoskeletal sarcomas. Nucl Med Commun 2006; 27:795–802 23. Kleis M, Daldrup-Link H, Matthay K, et al. Diagnostic value of PET/CT for the staging and restaging of pediatric tumors. Eur J Nucl Med Mol Imaging 2009; 36:23–36 24. Fuglø HM, Jørgensen SM, Loft A, Hovgaard D, Petersen MM. The diagnostic and prognostic value of 18F-FDG PET/CT in the initial assessment of high-grade bone and soft tissue sarcoma: a retrospective study of 89 patients. Eur J Nucl Med Mol Imaging 2012; 39:1416–1424 25. Cistaro A, Lopci E, Gastaldo L, Fania P, Brach Del Prever A, Fagioli F. The role of 18F-FDG
PET/CT in the metabolic characterization of lung nodules in pediatric patients with bone sarcoma. Pediatr Blood Cancer 2012; 59:1206–1210 26. Maki RG. Pediatric sarcomas occurring in adults. J Surg Oncol 2008; 97:360–368 27. Ulaner GA, Magnan H, Healey JH, Weber WA, Meyers PA. Is methylene diphosphonate bone scan necessary for initial staging of Ewing sarcoma if FDG PET/CT is performed? AJR 2014; 202:859–867 28. Benz MR, Czernin J, Tap WD, let al. FDGPET/CT imaging predicts histopathologic treatment responses after neoadjuvant therapy in adult primary bone sarcomas. Sarcoma 2010; 2010:143540 29. Byun BH, Kong CB, Lim I, et al.. Early response monitoring to neoadjuvant chemotherapy in osteosarcoma using sequential 18F-FDG PET/CT and MRI. Eur J Nucl Med Mol Imaging 2014; 41:1553–1562 30. Kager L, Zoubek A, Pötschger U, et al; Cooperative German-Austrian-Swiss Osteosarcoma Study Group. Primary metastatic osteosarcoma: presentation and outcome of patients treated on neoadjuvant Cooperative Osteosarcoma Study Group protocols. J Clin Oncol 2003; 21:2011–2018 31. Piperkova E, Mikhaeil M, Mousavi A, et al. Impact of PET and CT in PET/CT studies for staging and evaluating treatment response in bone and soft tissue sarcomas. Clin Nucl Med 2009; 34:146–150 32. Huda W, Vance A. Patient radiation doses from adult and pediatric CT. AJR 2007; 188:540–546 33. Fahey FH. Dosimetry of pediatric PET/CT. J Nucl Med 2009; 50:1483–1491 34. Buchbender C, Heusner TA, Lauenstein TC, Bockisch A, Antoch G. Oncologic PET/MRI. Part 2. Bone tumors, soft-tissue tumors, melanoma, and lymphoma. J Nucl Med 2012; 53:1244–1252
AJR:204, January 2015
