- Email: [email protected]
18F-fluorodeoxyglucose positron emission tomography–computed tomography for the evaluation of bone metastasis in patients with gastric cancer
Digestive and Liver Disease 45 (2013) 769–775
Contents lists available at SciVerse ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Oncology 18
F-fluorodeoxyglucose positron emission tomography–computed tomography for the evaluation of bone metastasis in patients with gastric cancer Dae Won Ma a,c , Jie-Hyun Kim a,c,∗ , Tae Joo Jeon b,c,∗∗ , Yong Chan Lee a , Mijin Yun b , Young Hoon Youn a,c , Hyojin Park a,c , Sang In Lee a,c a
Department of Internal Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea Department of Nuclear Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea c Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 4 October 2012 Accepted 13 February 2013 Available online 3 July 2013 Keywords: Bone metastasis 18 F-FDG PET–CT Gastric cancer WBBS
a b s t r a c t Background: The roles of positron emission tomography and bone scanning in identifying bone metastasis in gastric cancer are unclear. Aim: We compared the usefulness of positron emission tomography–computed tomography and scanning in detecting bone metastasis in gastric cancer. Methods: Data from 1485 patients diagnosed with gastric cancer who had undergone positron emission tomography–computed tomography and scanning were reviewed. Of 170 enrolled patients who were suspected of bone metastasis in either positron emission tomography or scanning, 81.2% were confirmed to have bone metastasis. Results: The sensitivity, specificity, and accuracy were 93.5%, 25.0%, and 80.6%, respectively, for positron emission tomography and 93.5%, 37.5%, and 82.9%, respectively, for scanning. 87.7% of patients with bone metastasis showed positive findings on two modalities. 15.0% of solitary bone metastases were positive on positron emission tomography only. Positron emission tomography was superior to scanning for the detection of synchronous bone metastasis, but the two modalities were similar for the detection of metachronous bone metastasis. The concordance rate of response assessment after treatment between two modalities was moderate. Conclusions: Positron emission tomography–computed tomography may be more effective for the diagnosis of bone metastasis in the initial staging workup. Conversely, bone scanning and positron emission tomography–computed tomography may be similarly effective for the detection of metachronous bone metastasis. © 2013 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Gastric cancer can metastasize to various sites at the time of diagnosis or during treatment. The most common site of metastasis in Asians is the peritoneum [1]. Other metastatic sites are the liver, lung, and bone. Evaluating the presence or absence of metastasis is important for the establishment of treatment plans for gastric cancer, which occurs at a high rate, and to predict
∗ Corresponding author at: Department of Internal Medicine, Yonsei University College of Medicine, 211 Eonjuro, Gangnam-gu, Seoul 135-720, Republic of Korea. Tel.: +82 2 2019 3505; fax: +82 2 3463 3882. ∗∗ Corresponding author at: Department of Nuclear Medicine, Yonsei University College of Medicine, 211 Eonjuro, Gangnam-gu, Seoul 135-720, Republic of Korea. Tel.: +82 2 2019 3741; fax: +82 2 3462 5472. E-mail addresses: [email protected] (J.-H. Kim), [email protected] (T.J. Jeon).
the prognosis. Few data exist regarding bone metastasis because this development is largely asymptomatic and radionuclide bone scintigraphy may not be a part of routine examination in the initial diagnostic workup. Additionally, the therapeutic approach for bone metastasis should be customized to facilitate risk stratification, providing the most appropriate therapy for each patient [2]. Many imaging methods have been shown to detect bone metastasis, including 18 F-fluorodeoxyglucose (FDG) positron emission tomography–computed tomography (PET–CT), whole-body bone scanning (WBBS), magnetic resonance imaging (MRI), and CT. The role of each imaging method in the detection of bone metastasis may vary according to the primary malignancy. PET–CT was found to be more useful than WBBS in detecting bone metastasis in patients with lung cancer [3]. However, PET–CT was not more accurate than WBBS in patients with head and neck cancer [4]. Previous
1590-8658/$36.00 © 2013 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.dld.2013.02.009
770
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
studies have found that the sensitivity, specificity, and accuracy of PET for the detection of bone metastasis were 93%, 94%, and 93%, respectively, in lung cancer, and 30%, 82%, and 57%, respectively, in gastric cancer [3,5]. Bone metastasis of gastric cancer is less common than that of other solid malignancies, such as breast, lung, and prostate cancers. The reported incidence of bone metastasis of gastric cancer has ranged from 1% to 20% [6–8]; this incidence has ranged from 0.9% to 2.4% in Korean studies [2,9]. The incidence of bone metastasis in gastric cancer has been difficult to define precisely. Nevertheless, the diagnosis of bone metastasis is important for patients with gastric cancer because its presence is an independent poor prognostic factor for advanced gastric cancer and it has adverse effects, such as severe pain or fracture [10,11]. However, clinical strategies for detecting bone metastasis in gastric cancer using PET–CT or WBBS, such as initial tumour staging workup or tumour assessment after treatment, have not been well established. Thus, the aims of the present study were to compare the usefulness of PET–CT and WBBS in detecting bone metastasis and to suggest a clinical approach to the diagnosis of bone metastasis in gastric cancer. 2. Patients and methods 2.1. Patient selection The electronic medical records at Severance and Gangnam Severance hospitals were reviewed retrospectively to identify patients with gastric cancer who had undergone PET–CT and WBBS between January 2003 and August 2011. In total, 1485 patients were retrospectively enrolled in our study on the basis of the following criteria: (1) histologically confirmed gastric cancer; (2) exclusion of other malignancy; (3) positive PET–CT or WBBS findings for bone metastasis; and (4) ≤3-month interval between PET–CT and
WBBS. All enrolled patients were confirmed to have gastric cancer pathologically. We excluded patients with coexistent malignancies, such as lung, colon, or ovarian cancer. All included patients showed positive findings for bone metastasis in one or more studies, and only patients who had undergone PET–CT and WBBS within 3 months were enrolled because the tumour can progress over time. As a result, 170 patients were retrospectively analysed. Bone metastasis was confirmed using the following criteria: (1) presence of increased radionuclide uptake throughout the skeleton on PET–CT or WBBS without reasonable explanation other than bone metastasis; (2) histologic examination; (3) compatible findings on CT or MRI; and (4) one or more equivocal lesions on PET–CT or WBBS with obvious progression on follow-up examinations [4]. If one of the criteria was met, bone metastasis was finally diagnosed. Thus, 138 patients (81.2%) were confirmed to have bone metastasis. Among them, 11 patients (8.0%) were diagnosed bone metastasis using follow-up examinations. There was no patient to be diagnosed bone metastasis using histology. All patients’ medical data were collected from the electronic medical records. Age, sex, operation history, pathologic information, and metastasis characteristics were reviewed retrospectively. Metastatic sites were categorized as the vertebrae, ribs, pelvic bones, long bones, sternum, scapulae, and skull. 2.2. Analysis of imaging modalities We divided patients into three groups by imaging modality findings (positive on PET–CT only, positive on WBBS only, and positive on both imaging modalities). In terms of analysis of PET–CT images, we evaluated both PET and CT images. At first, we checked PET images to detect hypermetabolic foci and then also reviewed bone setting CT images
Fig. 1. Case 1 shows the superiority of positron emission tomography–computed tomography to whole-body bone scans. (A) 18 F-fluorodeoxyglucose positron emission tomography–computed tomography reveals multiple hypermetabolic foci in the vertebrae, pelvic bone, and right proximal femur. (B) Although bone scintigraphy shows focal hot uptake in the right 7th rib and the lower margin of the third lumbar vertebral body, these findings can represent a traumatic reaction and degenerative change, respectively. That is, bone scan findings are negative [positron emission tomography, grade 1; whole-body bone scan, grade 3].
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
to detect the osteolytic or osteoblastic bone lesion, especially in marrow. Most metastatic cases showing osteolytic or no specific bone lesions in CT images revealed increased FDG uptake. While, some osteoblastic bone lesions showed only mild or no FDG uptake. Therefore, in the cases of controversy between PET and CT results, bony metastasis was diagnosed by the result from any one modality having positive finding. Positive metastatic findings of bone scan included definable photon defect only as well as abnormal hot uptake, and photon defect with marginal hot uptake of bone marrow. Improvement of bone metastasis with treatment was classified into three steps: improved, similar, and aggravated. Patients who had positive results on both imaging modalities were classified according to the superiority of the imaging modalities for the detection of bone metastasis. In these patients, exclusive of the same grade, the degree of detecting bone metastasis with the imaging modality was divided into three groups. Some examples are described below (Figs. 1–4). Grade 1 (excellent): one modality shows more metastatic bone lesions and greater detection ability than the other modality. Thus,
771
making a treatment plan and predicting a clinical course are possible. Grade 2 (fairly good): although one modality reveals more prominent pathologic lesions than the other modality, no influence on clinical course or treatment plan is expected. Grade 3 (questionable): one modality has a positive result for bone metastasis. However, other causes of bone lesions are possible and the other modality has more prominent results. WBBS was performed by intravenous administration of technetium-99m hydroxymethylene diphosphonate (740 MBq). Images were obtained after 4 h on a dual-head gamma camera. All patients underwent FDG-PET on a PET–CT unit (Biograph TruePoint 40; Siemens Healthcare, Erlangen, Germany). 120 kVp and 170 mAs for contrast enhanced CT scan. 120 kVp and 40 mAs for low dose CT (attenuation correction). Patients fasted for at least 6 h before the scan and had serum glucose levels <140 mg/dL. The injection dose of FDG was 5.5 MBq/kg body weight, with an intravenous administration route. All patients rested for 1 h in a dimly lit and quiet room before the scan.
Fig. 2. Case 2 shows the superiority of positron emission tomography–computed tomography to whole-body bone scans. (A) Strong focal fluorodeoxyglucose uptake in osteolytic bone lesions of the left sacral ala and left ischium are easily detected by positron emission tomography, whereas (B) bone scintigraphy reveals no significant abnormal focal bone uptake in these areas [positron emission tomography, grade 1; whole-body bone scan, grade 3].
772
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
Fig. 3. Case 3 shows the superiority of positron emission tomography–computed tomography to whole-body bone scans. (A) 18 F-fluorodeoxyglucose positron emission tomography–computed tomography reveals strong focal fluorodeoxyglucose uptake in the osteolytic bone lesion of the right acetabulum, while (B) bone scintigraphy shows asymmetric focal hot uptake that cannot clearly differentiate between bony metastasis and a degenerative change in the joint [positron emission tomography, grade 1; whole-body bone scan, grade 2].
Two experienced nuclear medicine physicians reviewed PET–CT and WBBS images. They achieved consensus on differences of opinion by mutual consultation.
3. Results
2.3. Statistical analysis
Seventy patients (52.9%) were male and the mean age at diagnosis of bone metastasis was 57.1 years. The mean interval between the diagnosis of gastric cancer and skeletal metastasis was 15.4 months. Undifferentiated histology by Japanese classification was demonstrated in 80.4% of patients with skeletal metastasis. Additionally, 49.3% of patients had evidence of bone metastasis at the time of gastric cancer diagnosis, which was categorized as synchronous bone metastasis. Conversely, 50.7% of patients showed evidence of bone metastasis during treatment, which was categorized as metachronous bone metastasis. Most patients (85.5%) had multiple bone metastases, and locations of bone metastases were
The sensitivity, specificity, positive and negative predictive values, and accuracy of the imaging modalities were calculated. Pearson’s chi-squared test and the Kruskal–Wallis test were used to compare modalities. A P-value <0.05 was deemed to indicate statistical significance. The kappa statistic () was used to determine the level of agreement between follow-up PET–CT and WBBS findings and to compare imaging method and overall tumour responses. The SPSS software (ver. 17.0; SPSS Inc., Chicago, IL, USA) was used for statistical analysis.
3.1. Baseline characteristics of patients
Fig. 4. Case 4 shows the superiority of whole-body bone scans to positron emission tomography–computed tomography. (A) Clearly defined focal hot uptake involving the fifth thoracic vertebral body and bilateral sacroiliac joints on bone scintigraphy suggest multiple bony metastases; however, (B) positron emission tomography shows no hypermetabolic focus in these areas, where a bone-setting computed tomographic image reveals multiple osteoblastic foci [positron emission tomography, grade 2; whole-body bone scan, grade 1].
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
similar to those reported in other studies. The common sites for bone metastasis were the vertebrae, pelvic bones, and ribs, in that order (data not shown). 3.2. Efficiency of imaging modalities PET–CT showed 93.5% sensitivity, 25.0% specificity, 84.3% positive predictive value, 47.1% negative predictive value, and 80.6% accuracy. One hundred fifty-three patients had positive findings on PET–CT and 129 of these patients were confirmed to have bone metastasis. Seventeen patients showed negative findings on PET–CT, but nine of these patients were confirmed to have bone metastasis. WBBS showed 93.5% sensitivity, 37.5% specificity, 86.6% positive predictive value, 57.1% negative predictive value, and 82.9% accuracy. Twenty-one patients showed negative findings on WBBS, but nine of these patients were confirmed to have bone metastasis. The sensitivity of WBBS was equal to that of PET–CT, but the specificity of WBBS was slightly higher than that of PET–CT. Only 21 patients had undergone PET–CT and WBBS more than twice after treatment, and we analysed initial and follow-up findings from both modalities. The concordance rate of tumour assessment after treatment between PET–CT and WBBS was analysed. The index between PET–CT and WBBS was 0.553, suggesting moderate agreement between findings from the two modalities. 3.3. Comparisons according to imaging modality One hundred thirty-eight patients with gastric cancer were confirmed to have bone metastasis. We classified these patients into three groups by imaging modality. Eight patients (5.8%) had a positive result for bone metastasis on PET–CT only, nine patients (6.5%) had a positive result for bone metastasis on WBBS only, and 121 patients (87.7%) had positive results on both modalities. We analysed these three groups in terms of the timing of diagnosis of bone metastasis and number of metastases (Supplementary Table 1). No significant difference was found among groups. Most patients (91.2%) who were diagnosed with bone metastasis at the initial assessment had positive results for bone metastasis on both modalities, and a similar result was shown in patients who were diagnosed with bone metastasis during follow-up (84.3%; P = 0.462). Three cases (15%) of solitary bone metastasis were positive on PET–CT only and nine cases (7.6%) of multiple bone metastasis were positive on WBBS only, although the results were not statistically significant (P = 0.084). Among 121 patients who had positive results on both imaging modalities, PET–CT was superior to WBBS in 15 cases, WBBS was superior to PET–CT in 7 cases, and the two modalities performed similarly for the detection of bone metastasis in 99 cases. Each patient was given a grade reflecting the superiority of the imaging modality for the detection of bone metastasis (Table 1). Three groups were analysed according to the timing of diagnosis with bone metastasis and number of bone metastases. Nine patients with synchronous metastasis were categorized as grade 1 on PET–CT, but no patient with synchronous metastasis was
773
Table 1 Imaging grade for analysis of superiority between positron emission tomography–computed tomography and whole-body bone scan (a n = 22). Grade
PET–CT (n, %)
WBBS (n, %)
1. Excellent 2. Fairly good 3. Questionable
15 (68.2) 5 (22.7) 2 (9.1)
7 (31.8) 5 (22.7) 10 (45.5)
P = 0.016. Abbreviations: PET–CT, positron emission tomography–computed tomography; WBBS, whole-body bone scan. a Cases with the same grade on both modalities (n = 99) were excluded.
categorized as grade 1 on WBBS. However, seven patients with metachronous metastasis were categorized as grade 1 on WBBS (P = 0.018; Table 2). 4. Discussion WBBS has been recognized as the most useful modality to evaluate the status of bone metastasis. PET is another modality that can be used to detect distant metastases, including bone metastasis. PET–CT has been used successfully to diagnose distant metastases in various malignancies. However, PET has not been found to be useful for detecting bone metastasis of gastric cancer. Yoshioka et al. [11] investigated the diagnostic accuracy of PET for advanced, metastatic, or recurrent gastric cancer. All PET images were interpreted visually, and tracer uptake was quantitated as the standardized uptake value (SUV). The authors found significantly higher SUVs for primary gastric lesions than for metastatic bone lesions, and thus suggested that PET may be a useful diagnostic modality for advanced, metastatic, or recurrent gastric cancer but not for detecting bone metastasis [11]. Another study reported that a comprehensive diagnosis based on a combination of PET, MRI, CT, WBBS, and clinical findings seemed necessary to confirm bone metastasis in gastric cancer [12]. No published study has investigated the efficacy of WBBS and PET–CT for detecting bone metastasis of gastric cancer, and no clinical algorithm has been reported. Thus, we aimed to analyse these issues in the present study. Our histologic data showed that nearly all gastric cancers (80.4%) were poorly differentiated and characterized as signet ring cell carcinoma, similar to the findings of previous studies [2,9,13]. The mean interval between diagnosis of gastric cancer and confirmation of bone metastasis was 15.4 months, and nearly half of the patients (49.3%) had bone metastasis at the time of gastric cancer diagnosis. These results are similar to those of previous studies [2,7,14]. According to those studies, including ours, patients with bone metastasis in gastric cancer display rapidly growing tumour biology. Thus, the early and precise evaluation and diagnosis of bone metastasis are important. In the present study, most (85.5%) bone metastases occurred at multiple sites and were detected predominantly in the axial skeleton, similar to a previous report [15]. This finding is supported by the proposed metastatic pathways of gastric cancer to the bones through the vertebral venous system [16] or the thoracic duct [17].
Table 2 Comparison according to imaging modality superiority. PET–CT superior (n = 15) Synchronous Metachronous Solitary Multiple
9 (14.5) 6 (10.2) 2 (11.8) 13 (12.5)
WBBS superior (n = 7)
PET–CT and WBBS similar (n = 99)
P
0 (0.0) 7 (11.9)
53 (85.5) 46 (78.0)
0.018
0 (0.0) 7 (6.7)
15 (88.2) 84 (80.8)
0.535
Abbreviations: PET–CT, positron emission tomography–computed tomography; WBBS, whole-body bone scan.
774
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
In the present study, PET–CT and WBBS showed similar sensitivity, specificity, and accuracy for the detection of bone metastasis in gastric cancer. However, we enrolled patients with suspected bone metastasis and with a positive result for bone metastasis on one or more imaging modalities. Thus, the sensitivity specificity, and accuracy of both studies could not but being calculated within limits. The true negative value may be extremely low; for this reason, the specificities of these modalities were lower in our study than in other studies. To compare the usefulness of PET–CT and WBBS for the detection of bone metastasis in gastric cancer, we analysed the characteristics of metastatic bone lesions with positive PET–CT findings only, positive WBBS findings only, and positivity on both modalities. No significant difference was found among these groups in the timing of bone metastasis occurrence or the number of bone metastases. Despite the lack of statistical significance, our results indicated that WBBS alone missed solitary bone metastasis in 15% of patients because this imaging modality sometimes exhibits the disadvantage of low specificity and produces false-positive results due to uptake by benign lesions, such as osteoarthritis, fractures, and inflammation. Using the semiquantitative parameter of SUV, Hur et al. [18] found that PET improved the diagnostic ability to differentiate between single bone metastases and benign lesions. Thus, PET–CT can be more helpful than WBBS for detecting a single bone metastasis. Additionally, PET–CT may diagnose bone metastasis earlier than WBBS. Single-photon emission computed tomography (SPECT) or SPECT/CT were not performed in our study. However, SPECT/CT could have improved the performance of WBBS in characterizing bone uptake. Although PET–CT and WBBS showed similar sensitivity for the detection of bone metastasis, we scored the imaging grades in positive lesions for both modalities as excellent, fairly good, or questionable to define the relative superiority of PET–CT and WBBS. The results showed that PET–CT detected twice the number of grade 1 lesions in comparison with WBBS. On cross-sectional analysis, PET–CT, which is generally used in patients with cancer for detecting distant metastases and recurrent disease, showed superiority in 14.5% of synchronous bone metastases [5]. Thus, PET–CT may be the preferred modality for detecting bone metastasis of gastric cancer at the initial staging workup. In contrast, PET–CT and WBBS showed similar performance for the detection of metachronous bone metastasis, and both modalities may thus be considered for follow-up examination. Furthermore, all cases in the WBBS-superior group showed metachronous bone metastasis. Thus, WBBS may be used as a follow-up imaging modality rather than for the initial staging workup. Biologic properties may account for the differences in radiologic characteristics of metastatic bone lesions between the imaging modalities. No definite practical guidelines exist to assess bone metastasis after treatment in patients with gastric cancer. Glucose metabolism of cancer cells decreases according to the level of treatment response. Thus, some studies have reported that early treatment evaluation by PET predicted responders and non-responders [19,20]. Our data show moderate agreement between PET–CT and WBBS findings after treatment. Both imaging modalities can be used complementarily to assess treatment response in patients with bone metastasis. Our study had some limitations. The retrospective study design may have introduced selection bias in our data. And, one of the limitations was that PET–CT and WBBS were included in the criteria to confirm bone metastasis. We found lower specificities of the imaging modalities than reported in other studies. Because we enrolled patients with suspected bone metastasis, the true negative value was very low. The treatment methods applied to patients were not standardized, but variably included approaches such as surgery, chemotherapy, and radiotherapy.
Our study did not analyse cost effectiveness of PET–CT and WBBS for bone metastasis in gastric cancer. However, cost-effectiveness analysis of two image modalities for bone metastasis can be helpful. In conclusion, PET–CT and WBBS have similar sensitivity for detecting bone metastasis in gastric cancer. However, PET–CT was superior to WBBS for detecting synchronous bone metastasis. Thus, PET–CT may be more effective than WBBS at the initial staging workup. To detect metachronous bone metastasis, WBBS and PET–CT are similarly effective. For response evaluation after treatment, both imaging studies may be considered complementary. However, a large-scale study is necessary to validate our data. Conflict of interest statement None declared. Acknowledgement This research was supported by the Basic Science Research Programme through the National Research Foundation of Korea (NRF), funded by the Ministry of Education, Science, and Technology (2009-0070279). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.dld.2013.02.009. References [1] Yoo CH, Noh SH, Shin DW, et al. Recurrence following curative resection for gastric carcinoma. British Journal of Surgery 2000;87:236–42. [2] Park HS, Rha SY, Kim HS, et al. A prognostic model to predict clinical outcome in gastric cancer patients with bone metastasis. Oncology 2011;80:142–50. [3] Min JW, Um SW, Yim JJ, et al. The role of whole-body FDG PET/CT, Tc 99m MDP bone scintigraphy, and serum alkaline phosphatase in detecting bone metastasis in patients with newly diagnosed lung cancer. Journal of Korean Medical Science 2009;24:275–80. [4] Kim MR, Roh JL, Kim JS, et al. 18 F-fluorodeoxyglucose-positron emission tomography and bone scintigraphy for detecting bone metastases in patients with malignancies of the upper aerodigestive tract. Oral Oncology 2008;44:148–52. [5] Shimada H, Okazumi S, Koyama M, et al. Japanese Gastric Cancer Association Task Force for Research Promotion: clinical utility of (1)F-fluoro2-deoxyglucose positron emission tomography in gastric cancer. A systematic review of the literature. Gastric Cancer 2011;14:13–21. [6] Noda N, Sano T, Shirao K, et al. A case of bone marrow recurrence from gastric carcinoma after a nine-year disease-free interval. Japanese Journal of Clinical Oncology 1996;26:472–5. [7] Yoshikawa K, Kitaoka H. Bone metastasis of gastric cancer. Japanese Journal of Surgery 1983;13:173–6. [8] Nishidoi H, Koga S. Clinicopathological study of gastric cancer with bone metastasis. Gan To Kagaku Ryoho 1987;14:1717–22. [9] Ahn JB, Ha TK, Kwon SJ. Bone metastasis in gastric cancer patients. Journal of Gastric Cancer 2011;11:38–45. [10] Takenaka D, Ohno Y, Matsumoto K, et al. Detection of bone metastases in nonsmall cell lung cancer patients: comparison of whole-body diffusion-weighted imaging (DWI), whole-body MR imaging without and with DWI, whole-body FDG-PET/CT, and bone scintigraphy. Journal of Magnetic Resonance Imaging 2009;30:298–308. [11] Yoshioka T, Yamaguchi K, Kubota K, et al. Evaluation of 18 F-FDG PET in patients with advanced, metastatic, or recurrent gastric cancer. Journal of Nuclear Medicine 2003;44:690–9. [12] Nakai T, Okuyama C, Kubota T, et al. FDG-PET in a case of multiple bone metastases of gastric cancer. Annals of Nuclear Medicine 2005;19:51–4. [13] Seto M, Tonami N, Koizumi K, et al. Bone metastasis in gastric cancer—clinical evaluation of bone scintigrams. Kaku Igaku 1983;20:795–801. [14] Nakanishi H, Araki N, Kuratsu S, et al. Skeletal metastasis in patients with gastric cancer. Clinical Orthopaedics and Related Research 2004:208–12. [15] Choi CW, Lee DS, Chung JK, et al. Evaluation of bone metastases by Tc-99m MDP imaging in patients with stomach cancer. Clinical Nuclear Medicine 1995;20:310–4. [16] Batson OV. The function of the vertebral veins and their role in the spread of metastases, 1940. Clinical Orthopaedics and Related Research 1995: 4–9. [17] Kamiya T, Honda K, Kizaki M, et al. Clinicopathological studies on disseminated carcinomatosis of the bone marrow occurring through metastasis of gastric carcinoma. Gan No Rinsho 1985;31:819–26.
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775 [18] Hur J, Yoon CS, Ryu YH, et al. Accuracy of fluorodeoxyglucose-positron emission tomography for diagnosis of single bone metastasis: comparison with bone scintigraphy. Journal of Computer Assisted Tomography 2007;31: 812–9. [19] Di Fabio F, Pinto C, Rojas Llimpe FL, et al. The predictive value of 18 FFDG-PET early evaluation in patients with metastatic gastric adenocarcinoma
775
treated with chemotherapy plus cetuximab. Gastric Cancer 2007;10: 221–7. [20] Ott K, Fink U, Becker K, et al. Prediction of response to preoperative chemotherapy in gastric carcinoma by metabolic imaging: results of a prospective trial. Journal of Clinical Oncology 2003;21:4604–10.
Contents lists available at SciVerse ScienceDirect
Digestive and Liver Disease journal homepage: www.elsevier.com/locate/dld
Oncology 18
F-fluorodeoxyglucose positron emission tomography–computed tomography for the evaluation of bone metastasis in patients with gastric cancer Dae Won Ma a,c , Jie-Hyun Kim a,c,∗ , Tae Joo Jeon b,c,∗∗ , Yong Chan Lee a , Mijin Yun b , Young Hoon Youn a,c , Hyojin Park a,c , Sang In Lee a,c a
Department of Internal Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea Department of Nuclear Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea c Gangnam Severance Hospital, Yonsei University College of Medicine, Seoul, Republic of Korea b
a r t i c l e
i n f o
Article history: Received 4 October 2012 Accepted 13 February 2013 Available online 3 July 2013 Keywords: Bone metastasis 18 F-FDG PET–CT Gastric cancer WBBS
a b s t r a c t Background: The roles of positron emission tomography and bone scanning in identifying bone metastasis in gastric cancer are unclear. Aim: We compared the usefulness of positron emission tomography–computed tomography and scanning in detecting bone metastasis in gastric cancer. Methods: Data from 1485 patients diagnosed with gastric cancer who had undergone positron emission tomography–computed tomography and scanning were reviewed. Of 170 enrolled patients who were suspected of bone metastasis in either positron emission tomography or scanning, 81.2% were confirmed to have bone metastasis. Results: The sensitivity, specificity, and accuracy were 93.5%, 25.0%, and 80.6%, respectively, for positron emission tomography and 93.5%, 37.5%, and 82.9%, respectively, for scanning. 87.7% of patients with bone metastasis showed positive findings on two modalities. 15.0% of solitary bone metastases were positive on positron emission tomography only. Positron emission tomography was superior to scanning for the detection of synchronous bone metastasis, but the two modalities were similar for the detection of metachronous bone metastasis. The concordance rate of response assessment after treatment between two modalities was moderate. Conclusions: Positron emission tomography–computed tomography may be more effective for the diagnosis of bone metastasis in the initial staging workup. Conversely, bone scanning and positron emission tomography–computed tomography may be similarly effective for the detection of metachronous bone metastasis. © 2013 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved.
1. Introduction Gastric cancer can metastasize to various sites at the time of diagnosis or during treatment. The most common site of metastasis in Asians is the peritoneum [1]. Other metastatic sites are the liver, lung, and bone. Evaluating the presence or absence of metastasis is important for the establishment of treatment plans for gastric cancer, which occurs at a high rate, and to predict
∗ Corresponding author at: Department of Internal Medicine, Yonsei University College of Medicine, 211 Eonjuro, Gangnam-gu, Seoul 135-720, Republic of Korea. Tel.: +82 2 2019 3505; fax: +82 2 3463 3882. ∗∗ Corresponding author at: Department of Nuclear Medicine, Yonsei University College of Medicine, 211 Eonjuro, Gangnam-gu, Seoul 135-720, Republic of Korea. Tel.: +82 2 2019 3741; fax: +82 2 3462 5472. E-mail addresses: [email protected] (J.-H. Kim), [email protected] (T.J. Jeon).
the prognosis. Few data exist regarding bone metastasis because this development is largely asymptomatic and radionuclide bone scintigraphy may not be a part of routine examination in the initial diagnostic workup. Additionally, the therapeutic approach for bone metastasis should be customized to facilitate risk stratification, providing the most appropriate therapy for each patient [2]. Many imaging methods have been shown to detect bone metastasis, including 18 F-fluorodeoxyglucose (FDG) positron emission tomography–computed tomography (PET–CT), whole-body bone scanning (WBBS), magnetic resonance imaging (MRI), and CT. The role of each imaging method in the detection of bone metastasis may vary according to the primary malignancy. PET–CT was found to be more useful than WBBS in detecting bone metastasis in patients with lung cancer [3]. However, PET–CT was not more accurate than WBBS in patients with head and neck cancer [4]. Previous
1590-8658/$36.00 © 2013 Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.dld.2013.02.009
770
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
studies have found that the sensitivity, specificity, and accuracy of PET for the detection of bone metastasis were 93%, 94%, and 93%, respectively, in lung cancer, and 30%, 82%, and 57%, respectively, in gastric cancer [3,5]. Bone metastasis of gastric cancer is less common than that of other solid malignancies, such as breast, lung, and prostate cancers. The reported incidence of bone metastasis of gastric cancer has ranged from 1% to 20% [6–8]; this incidence has ranged from 0.9% to 2.4% in Korean studies [2,9]. The incidence of bone metastasis in gastric cancer has been difficult to define precisely. Nevertheless, the diagnosis of bone metastasis is important for patients with gastric cancer because its presence is an independent poor prognostic factor for advanced gastric cancer and it has adverse effects, such as severe pain or fracture [10,11]. However, clinical strategies for detecting bone metastasis in gastric cancer using PET–CT or WBBS, such as initial tumour staging workup or tumour assessment after treatment, have not been well established. Thus, the aims of the present study were to compare the usefulness of PET–CT and WBBS in detecting bone metastasis and to suggest a clinical approach to the diagnosis of bone metastasis in gastric cancer. 2. Patients and methods 2.1. Patient selection The electronic medical records at Severance and Gangnam Severance hospitals were reviewed retrospectively to identify patients with gastric cancer who had undergone PET–CT and WBBS between January 2003 and August 2011. In total, 1485 patients were retrospectively enrolled in our study on the basis of the following criteria: (1) histologically confirmed gastric cancer; (2) exclusion of other malignancy; (3) positive PET–CT or WBBS findings for bone metastasis; and (4) ≤3-month interval between PET–CT and
WBBS. All enrolled patients were confirmed to have gastric cancer pathologically. We excluded patients with coexistent malignancies, such as lung, colon, or ovarian cancer. All included patients showed positive findings for bone metastasis in one or more studies, and only patients who had undergone PET–CT and WBBS within 3 months were enrolled because the tumour can progress over time. As a result, 170 patients were retrospectively analysed. Bone metastasis was confirmed using the following criteria: (1) presence of increased radionuclide uptake throughout the skeleton on PET–CT or WBBS without reasonable explanation other than bone metastasis; (2) histologic examination; (3) compatible findings on CT or MRI; and (4) one or more equivocal lesions on PET–CT or WBBS with obvious progression on follow-up examinations [4]. If one of the criteria was met, bone metastasis was finally diagnosed. Thus, 138 patients (81.2%) were confirmed to have bone metastasis. Among them, 11 patients (8.0%) were diagnosed bone metastasis using follow-up examinations. There was no patient to be diagnosed bone metastasis using histology. All patients’ medical data were collected from the electronic medical records. Age, sex, operation history, pathologic information, and metastasis characteristics were reviewed retrospectively. Metastatic sites were categorized as the vertebrae, ribs, pelvic bones, long bones, sternum, scapulae, and skull. 2.2. Analysis of imaging modalities We divided patients into three groups by imaging modality findings (positive on PET–CT only, positive on WBBS only, and positive on both imaging modalities). In terms of analysis of PET–CT images, we evaluated both PET and CT images. At first, we checked PET images to detect hypermetabolic foci and then also reviewed bone setting CT images
Fig. 1. Case 1 shows the superiority of positron emission tomography–computed tomography to whole-body bone scans. (A) 18 F-fluorodeoxyglucose positron emission tomography–computed tomography reveals multiple hypermetabolic foci in the vertebrae, pelvic bone, and right proximal femur. (B) Although bone scintigraphy shows focal hot uptake in the right 7th rib and the lower margin of the third lumbar vertebral body, these findings can represent a traumatic reaction and degenerative change, respectively. That is, bone scan findings are negative [positron emission tomography, grade 1; whole-body bone scan, grade 3].
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
to detect the osteolytic or osteoblastic bone lesion, especially in marrow. Most metastatic cases showing osteolytic or no specific bone lesions in CT images revealed increased FDG uptake. While, some osteoblastic bone lesions showed only mild or no FDG uptake. Therefore, in the cases of controversy between PET and CT results, bony metastasis was diagnosed by the result from any one modality having positive finding. Positive metastatic findings of bone scan included definable photon defect only as well as abnormal hot uptake, and photon defect with marginal hot uptake of bone marrow. Improvement of bone metastasis with treatment was classified into three steps: improved, similar, and aggravated. Patients who had positive results on both imaging modalities were classified according to the superiority of the imaging modalities for the detection of bone metastasis. In these patients, exclusive of the same grade, the degree of detecting bone metastasis with the imaging modality was divided into three groups. Some examples are described below (Figs. 1–4). Grade 1 (excellent): one modality shows more metastatic bone lesions and greater detection ability than the other modality. Thus,
771
making a treatment plan and predicting a clinical course are possible. Grade 2 (fairly good): although one modality reveals more prominent pathologic lesions than the other modality, no influence on clinical course or treatment plan is expected. Grade 3 (questionable): one modality has a positive result for bone metastasis. However, other causes of bone lesions are possible and the other modality has more prominent results. WBBS was performed by intravenous administration of technetium-99m hydroxymethylene diphosphonate (740 MBq). Images were obtained after 4 h on a dual-head gamma camera. All patients underwent FDG-PET on a PET–CT unit (Biograph TruePoint 40; Siemens Healthcare, Erlangen, Germany). 120 kVp and 170 mAs for contrast enhanced CT scan. 120 kVp and 40 mAs for low dose CT (attenuation correction). Patients fasted for at least 6 h before the scan and had serum glucose levels <140 mg/dL. The injection dose of FDG was 5.5 MBq/kg body weight, with an intravenous administration route. All patients rested for 1 h in a dimly lit and quiet room before the scan.
Fig. 2. Case 2 shows the superiority of positron emission tomography–computed tomography to whole-body bone scans. (A) Strong focal fluorodeoxyglucose uptake in osteolytic bone lesions of the left sacral ala and left ischium are easily detected by positron emission tomography, whereas (B) bone scintigraphy reveals no significant abnormal focal bone uptake in these areas [positron emission tomography, grade 1; whole-body bone scan, grade 3].
772
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
Fig. 3. Case 3 shows the superiority of positron emission tomography–computed tomography to whole-body bone scans. (A) 18 F-fluorodeoxyglucose positron emission tomography–computed tomography reveals strong focal fluorodeoxyglucose uptake in the osteolytic bone lesion of the right acetabulum, while (B) bone scintigraphy shows asymmetric focal hot uptake that cannot clearly differentiate between bony metastasis and a degenerative change in the joint [positron emission tomography, grade 1; whole-body bone scan, grade 2].
Two experienced nuclear medicine physicians reviewed PET–CT and WBBS images. They achieved consensus on differences of opinion by mutual consultation.
3. Results
2.3. Statistical analysis
Seventy patients (52.9%) were male and the mean age at diagnosis of bone metastasis was 57.1 years. The mean interval between the diagnosis of gastric cancer and skeletal metastasis was 15.4 months. Undifferentiated histology by Japanese classification was demonstrated in 80.4% of patients with skeletal metastasis. Additionally, 49.3% of patients had evidence of bone metastasis at the time of gastric cancer diagnosis, which was categorized as synchronous bone metastasis. Conversely, 50.7% of patients showed evidence of bone metastasis during treatment, which was categorized as metachronous bone metastasis. Most patients (85.5%) had multiple bone metastases, and locations of bone metastases were
The sensitivity, specificity, positive and negative predictive values, and accuracy of the imaging modalities were calculated. Pearson’s chi-squared test and the Kruskal–Wallis test were used to compare modalities. A P-value <0.05 was deemed to indicate statistical significance. The kappa statistic () was used to determine the level of agreement between follow-up PET–CT and WBBS findings and to compare imaging method and overall tumour responses. The SPSS software (ver. 17.0; SPSS Inc., Chicago, IL, USA) was used for statistical analysis.
3.1. Baseline characteristics of patients
Fig. 4. Case 4 shows the superiority of whole-body bone scans to positron emission tomography–computed tomography. (A) Clearly defined focal hot uptake involving the fifth thoracic vertebral body and bilateral sacroiliac joints on bone scintigraphy suggest multiple bony metastases; however, (B) positron emission tomography shows no hypermetabolic focus in these areas, where a bone-setting computed tomographic image reveals multiple osteoblastic foci [positron emission tomography, grade 2; whole-body bone scan, grade 1].
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
similar to those reported in other studies. The common sites for bone metastasis were the vertebrae, pelvic bones, and ribs, in that order (data not shown). 3.2. Efficiency of imaging modalities PET–CT showed 93.5% sensitivity, 25.0% specificity, 84.3% positive predictive value, 47.1% negative predictive value, and 80.6% accuracy. One hundred fifty-three patients had positive findings on PET–CT and 129 of these patients were confirmed to have bone metastasis. Seventeen patients showed negative findings on PET–CT, but nine of these patients were confirmed to have bone metastasis. WBBS showed 93.5% sensitivity, 37.5% specificity, 86.6% positive predictive value, 57.1% negative predictive value, and 82.9% accuracy. Twenty-one patients showed negative findings on WBBS, but nine of these patients were confirmed to have bone metastasis. The sensitivity of WBBS was equal to that of PET–CT, but the specificity of WBBS was slightly higher than that of PET–CT. Only 21 patients had undergone PET–CT and WBBS more than twice after treatment, and we analysed initial and follow-up findings from both modalities. The concordance rate of tumour assessment after treatment between PET–CT and WBBS was analysed. The index between PET–CT and WBBS was 0.553, suggesting moderate agreement between findings from the two modalities. 3.3. Comparisons according to imaging modality One hundred thirty-eight patients with gastric cancer were confirmed to have bone metastasis. We classified these patients into three groups by imaging modality. Eight patients (5.8%) had a positive result for bone metastasis on PET–CT only, nine patients (6.5%) had a positive result for bone metastasis on WBBS only, and 121 patients (87.7%) had positive results on both modalities. We analysed these three groups in terms of the timing of diagnosis of bone metastasis and number of metastases (Supplementary Table 1). No significant difference was found among groups. Most patients (91.2%) who were diagnosed with bone metastasis at the initial assessment had positive results for bone metastasis on both modalities, and a similar result was shown in patients who were diagnosed with bone metastasis during follow-up (84.3%; P = 0.462). Three cases (15%) of solitary bone metastasis were positive on PET–CT only and nine cases (7.6%) of multiple bone metastasis were positive on WBBS only, although the results were not statistically significant (P = 0.084). Among 121 patients who had positive results on both imaging modalities, PET–CT was superior to WBBS in 15 cases, WBBS was superior to PET–CT in 7 cases, and the two modalities performed similarly for the detection of bone metastasis in 99 cases. Each patient was given a grade reflecting the superiority of the imaging modality for the detection of bone metastasis (Table 1). Three groups were analysed according to the timing of diagnosis with bone metastasis and number of bone metastases. Nine patients with synchronous metastasis were categorized as grade 1 on PET–CT, but no patient with synchronous metastasis was
773
Table 1 Imaging grade for analysis of superiority between positron emission tomography–computed tomography and whole-body bone scan (a n = 22). Grade
PET–CT (n, %)
WBBS (n, %)
1. Excellent 2. Fairly good 3. Questionable
15 (68.2) 5 (22.7) 2 (9.1)
7 (31.8) 5 (22.7) 10 (45.5)
P = 0.016. Abbreviations: PET–CT, positron emission tomography–computed tomography; WBBS, whole-body bone scan. a Cases with the same grade on both modalities (n = 99) were excluded.
categorized as grade 1 on WBBS. However, seven patients with metachronous metastasis were categorized as grade 1 on WBBS (P = 0.018; Table 2). 4. Discussion WBBS has been recognized as the most useful modality to evaluate the status of bone metastasis. PET is another modality that can be used to detect distant metastases, including bone metastasis. PET–CT has been used successfully to diagnose distant metastases in various malignancies. However, PET has not been found to be useful for detecting bone metastasis of gastric cancer. Yoshioka et al. [11] investigated the diagnostic accuracy of PET for advanced, metastatic, or recurrent gastric cancer. All PET images were interpreted visually, and tracer uptake was quantitated as the standardized uptake value (SUV). The authors found significantly higher SUVs for primary gastric lesions than for metastatic bone lesions, and thus suggested that PET may be a useful diagnostic modality for advanced, metastatic, or recurrent gastric cancer but not for detecting bone metastasis [11]. Another study reported that a comprehensive diagnosis based on a combination of PET, MRI, CT, WBBS, and clinical findings seemed necessary to confirm bone metastasis in gastric cancer [12]. No published study has investigated the efficacy of WBBS and PET–CT for detecting bone metastasis of gastric cancer, and no clinical algorithm has been reported. Thus, we aimed to analyse these issues in the present study. Our histologic data showed that nearly all gastric cancers (80.4%) were poorly differentiated and characterized as signet ring cell carcinoma, similar to the findings of previous studies [2,9,13]. The mean interval between diagnosis of gastric cancer and confirmation of bone metastasis was 15.4 months, and nearly half of the patients (49.3%) had bone metastasis at the time of gastric cancer diagnosis. These results are similar to those of previous studies [2,7,14]. According to those studies, including ours, patients with bone metastasis in gastric cancer display rapidly growing tumour biology. Thus, the early and precise evaluation and diagnosis of bone metastasis are important. In the present study, most (85.5%) bone metastases occurred at multiple sites and were detected predominantly in the axial skeleton, similar to a previous report [15]. This finding is supported by the proposed metastatic pathways of gastric cancer to the bones through the vertebral venous system [16] or the thoracic duct [17].
Table 2 Comparison according to imaging modality superiority. PET–CT superior (n = 15) Synchronous Metachronous Solitary Multiple
9 (14.5) 6 (10.2) 2 (11.8) 13 (12.5)
WBBS superior (n = 7)
PET–CT and WBBS similar (n = 99)
P
0 (0.0) 7 (11.9)
53 (85.5) 46 (78.0)
0.018
0 (0.0) 7 (6.7)
15 (88.2) 84 (80.8)
0.535
Abbreviations: PET–CT, positron emission tomography–computed tomography; WBBS, whole-body bone scan.
774
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775
In the present study, PET–CT and WBBS showed similar sensitivity, specificity, and accuracy for the detection of bone metastasis in gastric cancer. However, we enrolled patients with suspected bone metastasis and with a positive result for bone metastasis on one or more imaging modalities. Thus, the sensitivity specificity, and accuracy of both studies could not but being calculated within limits. The true negative value may be extremely low; for this reason, the specificities of these modalities were lower in our study than in other studies. To compare the usefulness of PET–CT and WBBS for the detection of bone metastasis in gastric cancer, we analysed the characteristics of metastatic bone lesions with positive PET–CT findings only, positive WBBS findings only, and positivity on both modalities. No significant difference was found among these groups in the timing of bone metastasis occurrence or the number of bone metastases. Despite the lack of statistical significance, our results indicated that WBBS alone missed solitary bone metastasis in 15% of patients because this imaging modality sometimes exhibits the disadvantage of low specificity and produces false-positive results due to uptake by benign lesions, such as osteoarthritis, fractures, and inflammation. Using the semiquantitative parameter of SUV, Hur et al. [18] found that PET improved the diagnostic ability to differentiate between single bone metastases and benign lesions. Thus, PET–CT can be more helpful than WBBS for detecting a single bone metastasis. Additionally, PET–CT may diagnose bone metastasis earlier than WBBS. Single-photon emission computed tomography (SPECT) or SPECT/CT were not performed in our study. However, SPECT/CT could have improved the performance of WBBS in characterizing bone uptake. Although PET–CT and WBBS showed similar sensitivity for the detection of bone metastasis, we scored the imaging grades in positive lesions for both modalities as excellent, fairly good, or questionable to define the relative superiority of PET–CT and WBBS. The results showed that PET–CT detected twice the number of grade 1 lesions in comparison with WBBS. On cross-sectional analysis, PET–CT, which is generally used in patients with cancer for detecting distant metastases and recurrent disease, showed superiority in 14.5% of synchronous bone metastases [5]. Thus, PET–CT may be the preferred modality for detecting bone metastasis of gastric cancer at the initial staging workup. In contrast, PET–CT and WBBS showed similar performance for the detection of metachronous bone metastasis, and both modalities may thus be considered for follow-up examination. Furthermore, all cases in the WBBS-superior group showed metachronous bone metastasis. Thus, WBBS may be used as a follow-up imaging modality rather than for the initial staging workup. Biologic properties may account for the differences in radiologic characteristics of metastatic bone lesions between the imaging modalities. No definite practical guidelines exist to assess bone metastasis after treatment in patients with gastric cancer. Glucose metabolism of cancer cells decreases according to the level of treatment response. Thus, some studies have reported that early treatment evaluation by PET predicted responders and non-responders [19,20]. Our data show moderate agreement between PET–CT and WBBS findings after treatment. Both imaging modalities can be used complementarily to assess treatment response in patients with bone metastasis. Our study had some limitations. The retrospective study design may have introduced selection bias in our data. And, one of the limitations was that PET–CT and WBBS were included in the criteria to confirm bone metastasis. We found lower specificities of the imaging modalities than reported in other studies. Because we enrolled patients with suspected bone metastasis, the true negative value was very low. The treatment methods applied to patients were not standardized, but variably included approaches such as surgery, chemotherapy, and radiotherapy.
Our study did not analyse cost effectiveness of PET–CT and WBBS for bone metastasis in gastric cancer. However, cost-effectiveness analysis of two image modalities for bone metastasis can be helpful. In conclusion, PET–CT and WBBS have similar sensitivity for detecting bone metastasis in gastric cancer. However, PET–CT was superior to WBBS for detecting synchronous bone metastasis. Thus, PET–CT may be more effective than WBBS at the initial staging workup. To detect metachronous bone metastasis, WBBS and PET–CT are similarly effective. For response evaluation after treatment, both imaging studies may be considered complementary. However, a large-scale study is necessary to validate our data. Conflict of interest statement None declared. Acknowledgement This research was supported by the Basic Science Research Programme through the National Research Foundation of Korea (NRF), funded by the Ministry of Education, Science, and Technology (2009-0070279). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.dld.2013.02.009. References [1] Yoo CH, Noh SH, Shin DW, et al. Recurrence following curative resection for gastric carcinoma. British Journal of Surgery 2000;87:236–42. [2] Park HS, Rha SY, Kim HS, et al. A prognostic model to predict clinical outcome in gastric cancer patients with bone metastasis. Oncology 2011;80:142–50. [3] Min JW, Um SW, Yim JJ, et al. The role of whole-body FDG PET/CT, Tc 99m MDP bone scintigraphy, and serum alkaline phosphatase in detecting bone metastasis in patients with newly diagnosed lung cancer. Journal of Korean Medical Science 2009;24:275–80. [4] Kim MR, Roh JL, Kim JS, et al. 18 F-fluorodeoxyglucose-positron emission tomography and bone scintigraphy for detecting bone metastases in patients with malignancies of the upper aerodigestive tract. Oral Oncology 2008;44:148–52. [5] Shimada H, Okazumi S, Koyama M, et al. Japanese Gastric Cancer Association Task Force for Research Promotion: clinical utility of (1)F-fluoro2-deoxyglucose positron emission tomography in gastric cancer. A systematic review of the literature. Gastric Cancer 2011;14:13–21. [6] Noda N, Sano T, Shirao K, et al. A case of bone marrow recurrence from gastric carcinoma after a nine-year disease-free interval. Japanese Journal of Clinical Oncology 1996;26:472–5. [7] Yoshikawa K, Kitaoka H. Bone metastasis of gastric cancer. Japanese Journal of Surgery 1983;13:173–6. [8] Nishidoi H, Koga S. Clinicopathological study of gastric cancer with bone metastasis. Gan To Kagaku Ryoho 1987;14:1717–22. [9] Ahn JB, Ha TK, Kwon SJ. Bone metastasis in gastric cancer patients. Journal of Gastric Cancer 2011;11:38–45. [10] Takenaka D, Ohno Y, Matsumoto K, et al. Detection of bone metastases in nonsmall cell lung cancer patients: comparison of whole-body diffusion-weighted imaging (DWI), whole-body MR imaging without and with DWI, whole-body FDG-PET/CT, and bone scintigraphy. Journal of Magnetic Resonance Imaging 2009;30:298–308. [11] Yoshioka T, Yamaguchi K, Kubota K, et al. Evaluation of 18 F-FDG PET in patients with advanced, metastatic, or recurrent gastric cancer. Journal of Nuclear Medicine 2003;44:690–9. [12] Nakai T, Okuyama C, Kubota T, et al. FDG-PET in a case of multiple bone metastases of gastric cancer. Annals of Nuclear Medicine 2005;19:51–4. [13] Seto M, Tonami N, Koizumi K, et al. Bone metastasis in gastric cancer—clinical evaluation of bone scintigrams. Kaku Igaku 1983;20:795–801. [14] Nakanishi H, Araki N, Kuratsu S, et al. Skeletal metastasis in patients with gastric cancer. Clinical Orthopaedics and Related Research 2004:208–12. [15] Choi CW, Lee DS, Chung JK, et al. Evaluation of bone metastases by Tc-99m MDP imaging in patients with stomach cancer. Clinical Nuclear Medicine 1995;20:310–4. [16] Batson OV. The function of the vertebral veins and their role in the spread of metastases, 1940. Clinical Orthopaedics and Related Research 1995: 4–9. [17] Kamiya T, Honda K, Kizaki M, et al. Clinicopathological studies on disseminated carcinomatosis of the bone marrow occurring through metastasis of gastric carcinoma. Gan No Rinsho 1985;31:819–26.
D.W. Ma et al. / Digestive and Liver Disease 45 (2013) 769–775 [18] Hur J, Yoon CS, Ryu YH, et al. Accuracy of fluorodeoxyglucose-positron emission tomography for diagnosis of single bone metastasis: comparison with bone scintigraphy. Journal of Computer Assisted Tomography 2007;31: 812–9. [19] Di Fabio F, Pinto C, Rojas Llimpe FL, et al. The predictive value of 18 FFDG-PET early evaluation in patients with metastatic gastric adenocarcinoma
775
treated with chemotherapy plus cetuximab. Gastric Cancer 2007;10: 221–7. [20] Ott K, Fink U, Becker K, et al. Prediction of response to preoperative chemotherapy in gastric carcinoma by metabolic imaging: results of a prospective trial. Journal of Clinical Oncology 2003;21:4604–10.