detecting distant organ (M1b) and distant lymph node metastasis .... nonregional lymph node metastases, one was localized in the celiac region (M1a) and two.
7
PITFALLS OF POSITIVE FINDINGS IN STAGING ESOPHAGEAL CANCER WITH 18F-FLUORODEOXYGLUCOSE POSITRON EMISSION TOMOGRAPHY HL van Westreenen1 PAM Heeren1 PL Jager2 HM van Dullemen3 H Groen4 JThM Plukker1 Departments of Surgery1, Nuclear Medicine/PET center2, Gastroenterology3, Office for Medical Technology Assessment4 University Hospital Groningen, The Netherlands
Annals of Surgical Oncology 2003;10:1100-1105
CHAPTER 7
ABSTRACT Introduction: 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) is valuable in staging of esophageal cancer. However, FDG-PET may falsely upstage patients leading to incorrect exclusion from surgical treatment. This study was performed to determine the falsepositive rate and possible causes. Patients and Methods: The rate of false-positive lesions on FDG-PET was documented in 86 out of a group of 98 patients. Lesions were defined as false-positive when pathological examination was negative or as absence of tumor activity within 6 months of follow-up. To evaluate the influence of a learning curve on the false-positive rate, the PET scans were revised recently. Results: False-positive lesions were found in 13 patients (13 of 86; 15%). FDG-PET incorrectly revealed only locoregional node metastases in 5 patients in whom surgery with curative intent was performed. Ten lesions in the other 8 patients were classified as distant organ or as nonregional node metastases (M1a/1b). Finally, 5 patients upstaged to M1a/1b underwent a curative resection. The number of false-positive lesions decreased from 16 to 5 (6%) after revision. Conclusion: Proper interpretation of FDG-PET in staging esophageal cancer is impeded by false-positive results. Even after completion of the learning curve, positive FDG-PET findings still have to be confirmed by additional investigations.
92
False-Positive PET Results in Esophageal Cancer
INTRODUCTION The incidence of esophageal cancer is currently reported 3.2 per 100,000 persons. The incidence has shown a rising trend for 15 years. Surgical resection is still the only treatment with curative intent, and accurate staging is essential to select patients for esophagectomy. Traditionally, staging of esophageal cancer consists of computed tomography (CT), endoscopic ultrasonography (EUS), and sonographic examination of the neck.1 During the last decade, positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) as a tracer, has gained more and more acceptance as an additional staging tool in the preoperative work-up. Conventional staging modalities have the potential to give anatomic and morphological information about tumor extension, organ and lymph node metastases, whereas PET scanning identifies metabolically active cancer tissue based on a high glucose metabolism. FDG-PET has high sensitivity (70% to 90%) and specificity (90% to 100%) in detecting distant organ (M1b) and distant lymph node metastasis (M1a/1b).2-9 The main benefit of the high accuracy of FDG-PET in staging esophageal cancer is the reduction of useless surgery in patients with distant metastases. Whether whole-body FDG-PET can be used as a single-method imaging survey in esophageal cancer patients, however is questionable. The pitfall lies in the FDG-PET false-positive results, which may lead to improper upstaging of patients, and incorrect exclusion from curative treatment. Several studies have demonstrated that false-positive FDG-PET results are usually caused by inflammatory reactions.5,10,11 Other studies hypothesize inhomogeneous tracer uptake in the primary tumor site as a possible explanation for the occurrence of false-positive results.3,4 For most false-positive cases in staging of esophageal cancer in the literature, a clear reason is not described, nor has literature suggested ways to reduce false-positive FDG-PET results. As in other diagnostic areas, the learning curve in the rating of PET images may also play an important role in the accuracy of staging with FDGPET, but this has not yet been investigated in staging of esophageal cancer. The aim of this study was to document the false-positive rate in staging esophageal cancer with FDG-PET and to study potential causes of the false-positive results and their impact on clinical management. The role of a learning curve in reducing false-positive FDGPET results was investigated as well.
MATERIALS AND METHODS Patients A retrospective review was performed on the records of 98 consecutive patients with a biopsy-proven malignancy of the esophagus. All patients were staged with CT thorax and abdomen, EUS, and sonography of the neck and diagnosed between January 1996 and March 2002. In addition all patients underwent FDG-PET. Treatment of all patients was based on the conventional staging or confirmed FDG-PET findings. None of these patients received neoadjuvant treatment. The location of lesions was correlated to the pathological findings of resected specimens and biopsies obtained during surgery. If pathology was not available, lesions were verified by at least 6 months of follow-up time.12,13 Hotspots were defined as false
93
CHAPTER 7
Table 1. Characteristics of Included Patients (n=86) No. of patients (%) Sex Male
72 (84)
Female
14 (16)
Age*
21-78 (61)
Localization Gastroesophageal junction
43 (50)
Distal esophagus
38 (44)
Mid esophagus
7 (6)
Histopathology Adenocarcinoma
67 (78)
Squamous cell carcinoma
14 (16)
Other malignancies
5 (6)
*Range (median)
positive when pathological examination was negative for metastatic disease or when there was no clinical and radiological evidence for metastatic activity within 6 months of followup. Lesions seen by PET but not histologically confirmed were considered to be true-positive if clinical findings appeared at the site identified by FDG-PET within the 6 months of follow-up. An exclusion criterion was the absence of the standard reference. Based on this criterion, 86 patients (72 male and 14 female) were included (Table 1). Twelve patients (8 male/4 female) were excluded because the follow-up was shorter than 6 months. One patient died postoperatively; all other patients had an advanced stage of disease. Positron emission tomography FDG was produced according to the robotic method described by Hamacher et al14 with a radiochemical purity over 98%. An ECAT 951/31 or an ECAT HR+ positron camera (Siemens/ CTI, Knoxville, KY, USA) was used for data acquisition. The ECAT 951/31 acquires 31 planes over 10.9 cm, and the HR+ camera acquires 63 planes over a 15.8 cm axial field of view. All patients fasted for at least 4 hours before 400-580 MBq FDG was administered intravenously into an antecubital vein. Data acquisition started 90 minutes after injection in whole body mode, for 5 minutes per bed position from the skull to the knees. Transmission imaging was obtained during 3 minutes per bed position for attenuation correction. Images were reconstructed using an iterative reconstruction technique and were read from computer monitors. All PET-scans were originally interpreted by one of two nuclear medicine physicians
94
False-Positive PET Results in Esophageal Cancer
without knowledge of the CT findings nor EUS data and pathologic information. All hotspots of a non-physiologic FDG uptake were considered indicative for metastatic disease. Differentiation between locoregional and distant lymph node involvement was based on the location of the hotspot and the distance to the primary tumor. Tumors were classified as N0, N1, M1a or M1b by translation of positive PET findings to the current tumor, node, metastasis staging system of the International Union Against Cancer (UICC). All positive PET scans were recently revised by an experienced nuclear physician (PLJ) to investigate the effect of the learning curve as a possible cause of false-positive results. The nuclear physician was blinded for all clinical information and previous staging results and gave a classification according to the tumor, node, metastasis system. To compare the reduction of false-positive results after revision to the initial interpretation a paired-samples t-test was used. A value of P < 0.05 was considered significant.
RESULTS After comparing initial FDG-PET findings with pathological and/or follow-up results, we identified 13 (15%) patients with false-positive FDG-PET scans with a total number of 16 false-positive foci of metabolic lesions (Table 2). Three patients (6,7, and 13) had two falsepositive lesions in different locations. In five patients (1-5), FDG-PET showed foci suspected for locoregional metastases leading to incorrect N1 upstaging. Patient 6 was upstaged to
Table 2. Details of 16 False-Positive FDG-PET Results in 13 Patients Type of additional Patient
Localization
Upstaged
investigations
Results
Treatment
Reference
1.
Locoregional
N1
-
-
Resection
Pathology
2.
Locoregional
N1
-
-
Resection
Pathology
3.
Locoregional
N1
-
-
Resection
Pathology
4.
Locoregional
N1
-
-
Resection
Pathology
5.
Locoregional
N1
-
-
Resection
Pathology
6.
Locoregional
N1
-
-
Resection
Pathology
Cervical region
M1b
Sonography
Negative
7.
Celiac region
M1a
-
-
Explorative laparotomy
Follow-up Follow-up
8.
Liver
M1b
Sonography
Negative
Resection
Pathology
Rectum
M1b
Colonoscopy
Colitis
9.
Lung
M1b
-
-
Resection
Pathology Follow-up
10.
Rectum
M1b
Colonoscopy
Adenoma
Explorative laparotomy
Pathology
11.
Ribs
M1b
Bone scan
Negative
Resection
Follow-up
12.
Brain
M1b
-
-
Palliation
Follow-up
13.
Cheek
M1b
Dental tomography
Negative
Resection
Follow-up
Cervical region
M1b
Sonography
Negative
Follow-up
95
CHAPTER 7
N1 and M1b disease because of both a locoregional (N1) and a distant metastasis (M1b) were present on FDG-PET. Seven patients (7-13) with FDG-PET findings suspected for distant organ metastases or nonregional lymph node metastases were incorrectly upstaged to M1a or M1b. Only one of them (12) was confirmed to M1a/1b by conventional staging methods. Therefore, the FDG-PET findings were disregarded and the remaining patients underwent surgery with curative intent. Localization of false-positive lesions Seven false-positive lesions on FDG-PET were suspected for distant organ metastases (M1b). Of these lesions, one was localized in the liver, one in a lung, two in the rectum, one in a rib, one in the mandible, and one in the brain. Of the three lesions misinterpreted for nonregional lymph node metastases, one was localized in the celiac region (M1a) and two in the cervical region (M1b). Additional investigations Several additional investigations were performed to investigate the positive FDG-PET results. Lesions classified as locoregional lymph node metastases (N1) were not additionally investigated, because the presence of metastatic spread in locoregional nodes is not a contra indication for curative intended surgery. The lesions in the rectum of two patients were evaluated with colonoscopy and revealed colitis and an adenoma in the other patient, which was resected endoscopically. A maxillofacial surgeon examined the patient with a lesion in the mandible and dental tomography did not show any sign of malignancy. Bone scintigraphy to objectify the rib lesion could not confirm malignancy. The lesions in the liver and cervical region were examined sonographically, which revealed no abnormalities. Clinical management The five patients (1-5) who were upstaged to N1 only were enrolled for surgery, and they all had a curative resection. The resected specimens, however, did not reveal malignant lymph nodes. The patient (12) suspected of brain metastases on FDG-PET had other confirmed metastases at conventional staging methods and was restrained from surgical therapy. Two patients (7 and 10) upstaged to M1a and M1b underwent surgery, but their tumor was not curatively resectable, because of distant metastases neither visualized on conventional imaging methods or on FDG-PET. Therefore, these PET scans can be considered falsenegative as well. In the remaining 5 patients (6,8,10,11, and 13) upstaged to M1b based on FDG-PET results, a curative resection was performed. Follow-up Six patients (6,7,9, and 11-13) with a false-positive lesion, which could not be confirmed by histological or cytological examination, had a follow-up time of at least 6 months (range, 6-44 months). The patient (12) with false-positive brain lesions died after 6 months because of other confirmed metastases. The remaining patients had a follow-up time of at least 17 months (range, 17-44 months) without clinical tumor activity at the lesion. FDG-PET revision The revision of all positive FDG-PET resulted in a reduction from 16 to 5 false-positive lesions.
96
False-positives (%)
False-Positive PET Results in Esophageal Cancer
18 16 14 12 10 8 6 4
Initial
2
At revision
1996-1999
2000-2002
FIGURE 1
Percentage of false-positive cases in the period 1996-1999 (n = 43) and in the period 2000-2002 (n = 43).
The 6 locoregional metastases in the initial FDG-PET interpretation were all correctly staged as N0 at revision. The 10 distant metastases (M1a/1b) were reduced to 5. The suspected foci located in the celiac region, liver, brain, cheek, and one of the hotspots in the rectum were interpreted benign at revision. After revision, five patients were falsely upstaged to M1b disease (5.8%). In the first period from 1996 to 1999 the reduction of false-positive lesions was slightly higher (P = 0.16) than in the second period from 2000 to 2002. A false positive rate of 16.2% (7/43) in the first period reduced to 4.7% (2/43) after revision compared to the reduction from 14.1% (6/43) to 7.0% (3/43) in the second period (Figure 1).
DISCUSSION Surgical resection is still the only curative treatment in patients with esophageal cancer, but is associated with a considerable morbidity and even mortality. Proper selection for surgical treatment of patients is essential, because only a limited number of patients can be expected to profit from resection in advanced disease states. The role of whole body FDG-PET imaging in detecting metastatic disease that is already beyond cure has been demonstrated in several studies. Adding FDG-PET may alter the therapeutic management by correctly upstaging the classification in 10% to 15% of patients.3-5 However, interpretation of FDG-PET in staging of esophageal cancer, no less than in other disease states, is impeded by a substantial rate of false-positive results. The initial rate of false-positive findings in this study was 15%, consisting of 7% regional and 9% distant metastases. False-positive hotspots upstaged to M1a/1b disease in 5 of the 13 patients (38%) were additionally examined
97
CHAPTER 7
by several investigations revealing no abnormality. Therefore, these 5 patients were still considered resectable and finally underwent a curative resection. These findings emphasize the need to investigate positive FDG-PET results additionally before denying patients a chance of curative surgery. Possible allocation of adjuvant chemoradiotherapy in patients with N1 disease emphasizes the need to describe false-positive lesions leading to incorrect upstaging to N1 disease. The lesions falsely upstaged to N1 lesions may be caused by inhomogeneous tracer uptake in the primary tumor.3,4 In this study, five patients presented with false-positive N1 foci, and one patient was incorrectly upstaged to M1a, due to a lesion in de portal region. Inflammatory pulmonary disease may also lead to increased uptake in reactive regional lymph nodes.10 Shreve et al15 stated that abnormal accumulation of FDG in lymph nodes can be a consequence of spurious delivery of the tracer via lymphatic drainage, because the tracer extravasates into tissue drained by a regional lymph node group. Cervical lesions may also be caused by muscle activation leading to increased FDG metabolism or due to degenerative disease of the sternoclavicular joints.15 Verification of lesions by histological or cytological analysis is preferable, but not always obtainable nor necessary if other dedicated investigations are performed. When pathologic examination was not performed, we used a minimum follow-up time of 6 months.13 One patient died beyond 6 months without developing metastases in the lesion, and the other patients remained in the follow-up for more than 17 months. Therefore, we assume this as a reliable time period, because none of the patients in this study developed tumor activity in the lesion within 6 months. The causes of false-positive FDG-PET results in this study were only determined in the patients with lesions in the rectum due to adenoma and colitis. The rectal lesions may indicate peritoneal metastases of the esophageal carcinoma to the rectovesical or rectouterine pouch. On the other hand, a second primary tumor is also a reliable cause of lesions in the rectum. Therefore, the interpretation of these two lesions as false-positive is debatable. Other false-positive results are more difficult to elucidate. This study does not demonstrate more processes leading to increased FDG accumulation in nontumorous tissues, because the lack of cytological or histopathological examinations of all hotspots. Although FDG is a highly sensitive tumor tracer, it is not very specific as confirmed by the described false-positive cases. The rate of 7% false-positive local metastasis in this study is comparable to the 3% to 17% reported in the literature.3,4,10,16,17 However, the rate of incorrect upstaging to M1a/1b disease of 9% compares unfavorably to the range of 1% to 6% reported in literature.3,5,6,9 Since November 1995, FDG-PET has been used in staging esophageal cancer in our center. During this period of 7 years, experience has been gained in the interpretation of FDG-PET in about 100 esophageal tumors. As a result, the rate of falsepositive findings has decreased as shown in Figure 1. Revision of all positive FDG-PET scans by a blinded experienced nuclear medicine physician resulted in a lowering of the rate of falsepositive findings from 15% to 5.8%, especially relating to locoregional lesions. The reduction of false-positive FDG-PET results was higher in the first period, however, not significantly. Consequently, there seems to have been a learning curve influencing the interpretation of FDG-PET that will reduce the occurrence of false-positive results. Another factor to improve interpretation of FDG-PET findings will be the use of PET/
98
False-Positive PET Results in Esophageal Cancer
CT, especially of lesions located in the upper abdomen. This technique enables a precise mapping of increased FDG uptake to the anatomic background. However, no literature about this subject currently exists, and PET/CT is not yet available in our hospital. The increased glucose metabolism of malignant cells is the rationale behind FDG as the current most commonly used radiotracer in oncological PET studies.18 Several mechanisms have been proposed to account for FDG accumulation in tumors, including elevated levels of hexokinase and decreased activity of glucose-6-phosphatase. The increased glycolysis in malignant tumors with up-regulation of glucose transporter proteins is expressed by an enhancement of GLUT-1. GLUT-1 transporter is overexpressed in a wide variety of cancers including breast, lung cancer, colorectal, esophageal, and gastric adenocarcinoma.19 Misinterpretation of tumor staging with PET can also be found in inflammatory tissues, abscesses, autoimmune lesions, sarcoidosis, and some benign tumors. Focal hypermetabolic area due to high uptake of FDG in macrophages, lymphocytes, plasma cells, and neutrophils may render false-positive findings.11,20-23 To differentiate between malignant en benign lesions physiologic uptake of FDG is misleading. This physiology includes uptake in digestive tract, thyroid gland, skeletal muscle, myocardium, and bone marrow. The use of a semiquantitative measurement of standardized uptake value (SUV) in differentiating malignant tissue from benign on FDGPET is still questionable.24 Optimal SUV is computed for some tumors to classify a hotspot as malignant or benign but other studies demonstrated the opposite. The interpretation of FDG-PET in staging esophageal cancer will be improved by several factors leading to a decrease of false-positive findings. At first the awareness of physiologic and inflammatory processes resulting in FDG accumulation. Second, the interpreter must be informed in detail about both clinical condition and history of the patient at the moment of scanning. Finally, the experience of the nuclear physician will be a valuable factor influencing the false-positive rate. In conclusion, this study demonstrates the pitfalls of staging esophageal cancer with FDG-PET due to the occurrence of false-positive results. We should remember that FDG is not a tumor-specific substance, and that false-positive results may occur as a result of increased glucose metabolism in benign lesions. This study showed that PET still has to be used complementary to conventional staging methods. From these observations, it is clear that positive findings on FDG-PET must be confirmed by pathological examination, whenever possible, before denying patients from surgery with curative intent.
99
CHAPTER 7
REFERENCES 1. 2. 3.
4. 5. 6. 7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
100
Rice TW. Clinical staging of esophageal carcinoma. CT, EUS, and PET. Chest Surg Clin N Am 2000;10:471-85. Block MI, Patterson GA, Sundaresan RS, Bailey MS, Flanagan FL, Dehdashti F, Siegel BA, Cooper JD. Improvement in staging of esophageal cancer with the addition of positron emission tomography. Ann Thorac Surg 1997;64:770-6. Flamen P, Lerut A, Van Cutsem E, De Wever W, Peeters M, Stroobants S, Dupont P, Bormans G, Hiele M, de Leyn P, Van Raemdonck D, Coosemans W, Ectors N, Haustermans K, Mortelmans L. Utility of positron emission tomography for the staging of patients with potentially operable esophageal carcinoma. J Clin Oncol 2000;18:3202-10. Flanagan FL, Dehdashti F, Siegel BA, Trask DD, Sundaresan SR, Patterson GA, Cooper JD. Staging of esophageal cancer with 18F-fluorodeoxyglucose positron emission tomography. Am J Roentgenol 1997;168:417-24. Kole AC, Plukker JT, Nieweg OE, Vaalburg W. Positron emission tomography for staging of oesophageal and gastroesophageal malignancy. Br J Cancer 1998;78:521-7. Luketich JD, Schauer PR, Meltzer CC, Landreneau RJ, Urso GK, Townsend DW, Ferson PF, Keenan RJ, Belani CP. Role of positron emission tomography in staging esophageal cancer. Ann Thorac Surg 1997;64:765-9. Meltzer CC, Luketich JD, Friedman D, Charron M, Strollo D, Meehan M, Urso GK, Dachille MA, Townsend DW. Whole-body FDG positron emission tomographic imaging for staging esophageal cancer comparison with computed tomography. Clin Nucl Med 2000;25:882-7. Wallace MB, Nietert PJ, Earle C, Krasna MJ, Hawes RH, Hoffman BJ, Reed CE. An analysis of multiple staging management strategies for carcinoma of the esophagus: computed tomography, endoscopic ultrasound, positron emission tomography, and thoracoscopy/laparoscopy. Ann Thorac Surg 2002;74:1026-32. Wren SM, Stijns P, Srinivas S. Positron emission tomography in the initial staging of esophageal cancer. Arch Surg 2002;137:1001-6. Choi JY, Lee KH, Shim YM, Lee KS, Kim JJ, Kim SE, Kim BT. Improved detection of individual nodal involvement in squamous cell carcinoma of the esophagus by FDG PET. J Nucl Med 2000;41:80815. Zhuang H, Pourdehnad M, Lambright ES, Yamamoto AJ, Lanuti M, Li P, Mozley PD, Rossman MD, Albelda SM, Alavi A. Dual time point 18F-FDG PET imaging for differentiating malignant from inflammatory processes. J Nucl Med 2001;42:1412-7. Duarte PS, Zhuang H, Castellucci P, Alavi A. The receiver operating characteristic curve for the standard uptake value in a group of patients with bone marrow metastasis. Mol Imaging Biol 2002;4:157-60. Nguyen AT, Akhurst T, Larson SM, Coit DG, Brady MS. PET Scanning with (18)F 2-Fluoro-2-Deoxy-DGlucose (FDG) in Patients with Melanoma. Benefits and Limitations. Clin Positron Imaging 1999;2:938. Hamacher K, Coenen HH, Stocklin G. Efficient stereospecific synthesis of no-carrier-added 2-[18F]fluoro-2-deoxy-D-glucose using aminopolyether supported nucleophilic substitution. J Nucl Med 1986;27:235-8. Shreve PD, Anzai Y, Wahl RL. Pitfalls in oncologic diagnosis with FDG PET imaging: physiologic and benign variants. Radiographics 1999;19:61-77. Kato H, Kuwano H, Nakajima M, Miyazaki T, Yoshikawa M, Ojima H, Tsukada K, Oriuchi N, Inoue T, Endo K. Comparison between positron emission tomography and computed tomography in the use of the assessment of esophageal carcinoma. Cancer 2002;94:921-8. Kim K, Park SJ, Kim BT, Lee KS, Shim YM. Evaluation of lymph node metastases in squamous cell carcinoma of the esophagus with positron emission tomography. Ann Thorac Surg 2001;71:290-4. Czernin J, Phelps ME. Positron emission tomography scanning: current and future applications. Annu Rev Med 2002;53:89-112. Phay JE, Hussain HB, Moley JF. Strategy for identification of novel glucose transporter family members by using internet-based genomic databases. Surgery 2000;128:946-51. Ishimori T, Saga T, Mamede M, Kobayashi H, Higashi T, Nakamoto Y, Sato N, Konishi J. Increased
False-Positive PET Results in Esophageal Cancer
21. 22. 23. 24.
(18)F-FDG uptake in a model of inflammation: concanavalin A- mediated lymphocyte activation. J Nucl Med 2002;43:658-63. Kubota R, Yamada S, Kubota K, Ishiwata K, Tamahashi N, Ido T. Intratumoral distribution of fluorine18-fluorodeoxyglucose in vivo: high accumulation in macrophages and granulation tissues studied by microautoradiography. J Nucl Med 1992;33:1972-80. Strauss LG. Fluorine-18 deoxyglucose and false-positive results: a major problem in the diagnostics of oncological patients. Eur J Nucl Med 1996;23:1409-15. Yamada S, Kubota K, Kubota R, Ido T, Tamahashi N. High accumulation of fluorine-18fluorodeoxyglucose in turpentine- induced inflammatory tissue. J Nucl Med 1995;36:1301-6. Cremerius U, Wildberger JE, Borchers H, Zimny M, Jakse G, Gunther RW, Buell U. Does positron emission tomography using 18-fluoro-2-deoxyglucose improve clinical staging of testicular cancer?-Results of a study in 50 patients. Urology 1999;54:900-4.
101
