- Email: [email protected]
CT superior to serum CEA in detection of colorectal cancer and its recurrence
Clinical Imaging 37 (2013) 1094–1097
Contents lists available at ScienceDirect
Clinical Imaging journal homepage: http://www.clinicalimaging.org
18F-FDG PET/CT superior to serum CEA in detection of colorectal cancer and its recurrence☆ William Makis a,⁎, David Kurzencwyg b, Marc Hickeson b a b
Department of Diagnostic Imaging, Cross Cancer Institute, 11560 University Ave NW Edmonton, AB, T6G 1Z2, Canada Department of Nuclear Medicine, Royal Victoria Hospital, McGill University, 687 Pine Ave West, Floor M2, Montreal, Quebec, H3A 1A1, Canada
a r t i c l e
i n f o
Article history: Received 13 January 2013 Accepted 29 April 2013 Keywords: FDG PET/CT CEA Carcinoembryonic antigen Colorectal cancer
a b s t r a c t The aim of this study was to examine whether positron emission tomography (PET)/computed tomography (CT) can detect more cases of colorectal cancer (CRC) than serum carcinoembryonic antigen (CEA), both at initial staging and during surveillance for recurrence. A retrospective review of 639 CRC patients imaged with PET/CT was performed. PET/CT was superior to serum CEA in detecting CRC, identifying 2.5 times as many CRC at initial staging compared to serum CEA and 1.5 times as many CRC recurrences. The current guideline recommendations of utilizing PET/CT only in the context of a rising serum CEA will miss more than one third of all CRC recurrences. © 2013 Elsevier Inc. All rights reserved.
1. Introduction Since its introduction in 1965, serum carcinoembryonic antigen (CEA) has been an important tumor marker in the management of colorectal cancer (CRC) [1]. CEA does not play a significant role in the staging of CRC due to its low sensitivity [2]; however, serial CEA measurements are recommended by both the 2008 National Comprehensive Cancer Network (NCCN) practice guidelines as well as the 2005 American Society of Clinical Oncology (ASCO) guidelines for the surveillance for CRC recurrence [3]. Positron emission tomography (PET)/computed tomography (CT) is a hybrid imaging modality that takes advantage of the high rate of glycolysis in malignant tumor cells (PET), as well as precise anatomic correlation (CT). PET/CT has not been studied extensively in the initial staging of CRC mainly due to low reported sensitivity in detecting local lymph node involvement [4]. But recently, PET/CT has been reported to have a very high sensitivity and specificity in the detection of CRC recurrence. In a study of 68 patients with CRC recurrence, Chen reported a sensitivity of 94.6% and specificity of 83.3% with PET/CT [5]. Similar results were reported by Cohade [6]. The 2008 NCCN practice guidelines for colorectal cancer suggest performing a PET scan only in the context of an elevated serum CEA.
However, to our knowledge, a direct comparison between PET/CT results and serum CEA has not been done to justify this recommendation. The objective of this paper is to determine if PET/CT can detect more CRC recurrences than elevated serum CEA levels would suggest, and whether a revision of current guidelines for surveillance of CRC recurrence is needed.
2. Materials and methods 2.1. Study design and patient population A retrospective review of 639 patients referred for 18F-FDG PET/CT imaging to the McGill University Health Centre between September 2005 and October 2007 for initial staging of CRC or surveillance of CRC recurrence was performed. Of the 639 patients imaged, 328 had corresponding serum CEA measurements done within 4 months of the PET/CT scan. These were subsequently categorized into two groups: the initial staging group had 189 patients, and the surveillance of CRC recurrence group had 139 patients.
2.2. CEA measurement ☆ This manuscript or any figure or table in it has not been submitted to any publication previously. None of the authors have any financial or other relationships that might lead to a conflict of interest. The manuscript has been read and approved by all the authors, and the requirements for authorship have been met. Each author believes that the manuscript represents honest work. ⁎ Corresponding author. Tel.: +1 780 432 8760; fax: +1 780 432 8411. E-mail address: [email protected] (W. Makis). 0899-7071/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clinimag.2013.04.004
Between September 2005 and July 24, 2007, serum CEA measurements were done with the CEA-Roche Kit (normal values: 0–4.9). After July 24, 2007, serum CEA measurements were done with the Beckman Coulter Access Immunoassay system (Burlington, Ontario, Canada) (normal values: 0–3.4).
W. Makis et al. / Clinical Imaging 37 (2013) 1094–1097
2.3.
18
F-FDG PET/CT imaging
18
F-FDG PET/CT studies were performed on a hybrid PET/CT scanner (Discovery ST, General Electric Medical Systems, Waukesha, WI, USA), which combines a dedicated, full-ring PET scanner with a 16-slice CT scanner. Patients were required to fast for at least 6 h before the time of their appointment and waited in a quiet dark room the morning prior to their scan. Blood glucose levels were recorded immediately prior to 18F-FDG administration. If the serum glucose level was greater than 11.1 mmol/l (200 mg/dl), then the study was rescheduled. A volume of 400 ml of barium sulfate oral contrast was administered, and 0.22 mCi/kg of 18F-FDG was injected intravenously, up to a maximum dose of 20 mCi. Approximately 60 min following 18 F-FDG injection, CT and PET images were consecutively acquired from the base of the skull to the upper thighs, with additional images of the extremities acquired if needed. For the CT scan portion of the study, the settings were as follows: 140 kVp, 90–110 mA (depending on the body weight), a rotation time of 0.8 s, a table speed of 17 mm per gantry rotation, a pitch of 1.75:1, and a detector row configuration of 16×0.625 mm. For the PET portion of the study, a two-dimensional acquisition was performed, images were acquired using 4–5 min per bed position (depending on the body weight), and five to six bed positions were used (depending on the patient's height). The patient was allowed to breathe normally during the PET and CT acquisitions. Data obtained from the CT acquisition were used for attenuation correction and fusion with PET images. The PET data were reconstructed iteratively using the ordered subset expectation maximization software provided by the manufacturer (21 subsets, 2 iterations). PET attenuation corrected, PET non-attenuation corrected, CT, and PET/CT fusion images of the whole body were displayed in the axial, coronal, and sagittal planes and were reviewed on a dedicated workstation (Xeleris 2.0, GE Healthcare, Waukesha, WI, USA). PET data were also displayed in a rotating maximum intensity projection (MIP) image. 2.4. Interpretation and analysis of PET/CT images All PET/CT images were interpreted using visualization and semiquantitative analysis (SUVmax corrected for body weight) by two nuclear medicine physicians—who have a broad experience in clinical PET/CT and are familiar with normal variants, artifacts, and pitfalls in the interpretation of PET/CT studies—independently. PET/CT results were classified as positive or suspicious for neoplasm versus negative or not suspicious for neoplasm. 3. Results From the 189 patients referred for initial staging of CRC, PET/CT detected approximately 2.5 times (n=182) as many CRCs as CEA (n= 73) (Table 1). All 182 PET/CT-positive cases had histologically proven CRC. Seven patients had both negative PET/CT and CEA, and subsequent workup did not reveal any evidence of CRC in these patients. From the 139 patients referred for surveillance of CRC recurrence, PET/CT detected approximately 1.5 times (n=99) as many CRC recurrences as CEA (n=65). All 99 PET/CT positive cases had radiologic or histologic confirmation of CRC recurrence. Ten patients had a negative PET/CT with a positive CEA, eight of which (80%) did not show Table 1 PET/CT and CEA results for staging scans and surveillance scans of 328 CRC patients Staging scans (n=189)
CEA + CEA −
Surveillance scans (n=139)
PET +
PET −
PET +
PET −
73 109
0 7
65 34
10 30
1095
any evidence of disease recurrence in the subsequent 18 months of clinical and radiological follow-up. One patient had a brain metastasis, which was not detected by PET/CT as it was not in the image acquisition range; however, this finding was known from prior clinical and radiologic evaluation. Another patient had a negative PET/CT with a borderline-elevated CEA (3.6 with normal range of 0–3.4). However, the PET/CT became positive on a subsequent scan done 3 months after the initial negative scan. None of the PET/CT-negative and CEA-positive cases were mucinous adenocarcinomas histologically. 4. Discussion 4.1. PET/CT as initial diagnostic staging modality Serum CEA has a very low sensitivity in the diagnostic workup of CRCs. In a recent study of 425 CRC patients, only 42.6% of all diagnosed CRC patients had an elevated preoperative CEA [2]. In our series, PET/ CT detected 2.5 times as many CRC at initial staging as serum CEA (Fig. 1). These findings suggest that despite the shortcomings of its predecessor (PET), PET/CT's role in the initial staging of CRC warrants closer examination. 4.2. PET/CT in detection of CRC recurrence and metastases Recurrence of CRC occurs in about one third of patients within the first 2 years after colonic resection or rectal amputation and is most frequent in the area contiguous with the surgery. The ASCO 2005 guideline for stage II and III CRC patients recommends measuring serum CEA every 3 months and performing CT scans of the abdomen yearly [3]. However, CEA also has a low sensitivity in the detection of CRC recurrence. Moertel et al. reported a sensitivity of 59% and specificity of 84% for serum CEA in detecting CRC recurrence [7]. Recent attempts have been made to increase the sensitivity of serial CEA measurements by various mathematical manipulations, including calculating a CEA doubling time as well as a CEA half-time, with uncertain results [8]. However, even if CEA detects CRC recurrence, it does not provide any information about the location of the recurrence. CT also has suboptimal sensitivity and specificity in the detection of CRC recurrence, with one meta-analysis reporting a sensitivity of 79% and specificity of 73% [9]. Initial studies done with 18F-FDG PET validated its use in detection of hepatic metastases from CRC, but PET's role in local and regional disease recurrence was less clear. The modality proved to be problematic, as abnormalities could not always be anatomically localized and the rate of false positives was significant. The hybrid PET/CT imaging modality has been reported to significantly increase the sensitivity of PET in the detection of recurrence of CRC. The CT adds excellent anatomical resolution (the increased sensitivity is the result of CT's ability to detect small lung and liver metastases) and physiologic or inflammatory glucose uptake can usually be correctly interpreted on the basis of typical CT findings, which increases specificity. Votrubova et al. showed an increase in sensitivity and specificity from 80% and 69% for PET to 89% and 92% for PET/CT in detecting CRC recurrence in 84 patients [10]. Chen et al. found the sensitivity and specificity of PET/CT to be 94.6% and 83.3% in their 68 patient study [5]. Despite these excellent results, PET/CT is not yet considered an evidence-based tool in the diagnosis of CRC recurrence. 4.3. Positive PET/CT with negative CEA (for CRC recurrence) Our results showed that PET/CT detected 1.5 times as many CRC recurrences as CEA alone. Hence, any protocol or guideline recommendation of performing PET/CT only in the context of a positive CEA will miss a large number (34.3%) of patients who have CRC recurrence, with completely normal CEA levels, even on serial CEA measurements (Fig. 2).
1096
W. Makis et al. / Clinical Imaging 37 (2013) 1094–1097
Fig. 1. Staging PET/CT positive, CEA negative. A 58-year-old man imaged with a staging PET/CT preoperatively for colorectal carcinoma had an FDG-positive rectal mass and a liver metastasis (arrows), both confirmed on biopsy. Serum CEA level done 1 week prior to PET/CT was in the normal range (0.7; normal: 0–4.9). (A) Coronal CT, (B) PET, (C) PET/CT fusion, and (D) MIP images showing FDG-positive rectal mass and liver metastasis.
An argument could be made that PET/CT results should only be compared to CEA results when evaluating tumors that have elevated pretreatment CEA levels to begin with. Of our 39 surveillance PET scans that were PET positive and CEA negative, preoperative CEA levels were not available; however, 21 had serial CEA measurements done posttreatment. Of these, four (19%) became CEA positive with a lag time of 4–12 months following the positive PET/CT result. A similar result was observed by Kang et al., who followed CRC patients, all of whom had positive preoperative serum CEA levels. Following therapy, 69 of their patients had a positive PET/CT scan, and 20% of them had negative CEA levels [11]. PET/CT outperforms serum CEA, even when only looking at CEAproducing colorectal tumors. However, it is important to note that none of the current guidelines make a distinction between CEA- or non-CEA-producing colorectal tumors when determining whether a PET/CT should be done or not. This is another shortcoming of the current guidelines.
4.4. Negative PET/CT with positive CEA Ten of our 139 patients imaged for surveillance of CRC recurrence had a negative PET/CT with a positive CEA, of which 8 did not show any evidence of disease recurrence in the subsequent 18 months of clinical and radiologic follow-up, with CEA levels remaining stable. In the two cases that did show evidence of disease recurrence, one had a pathologically proven brain metastasis, which was not detected by PET/CT as it was not in the image acquisition range. The other case had a borderline-elevated CEA, which became significantly elevated as well as PET/CT positive just 3 months later. None of the PET/CT-negative and CEA-positive cases were mucinous adenocarcinomas histologically. Similar results were reported by Kang et al. from their 171 patients; only three had a negative PET with a positive CEA. None of these three patients had evidence of disease recurrence after further investigation [11]. These results suggest that, in most cases where the CEA is elevated but PET/CT is negative, it is the PET/CT that gives the
Fig. 2. Surveillance PET/CT positive, CEA negative. A 64-year-old man previously diagnosed with colorectal carcinoma, treated surgically 3 years prior, had a follow-up PET/CT after an abdominal CT showed a growing adrenal mass. The PET/CT showed bilateral FDG-positive lung metastases and a right adrenal metastasis (arrows). Serum CEA level done 2 months following PET/CT was in the normal range (3.7; normal: 0–4.9). (A) Coronal CT, (B) PET, (C) PET/CT fusion, and (D) MIP images showing FDG-positive bilateral lung metastases and a right adrenal metastasis.
W. Makis et al. / Clinical Imaging 37 (2013) 1094–1097
correct diagnosis. In our study, serum CEA detected disease that was not detected by PET/CT in only two cases; however, the clinical management in these two cases was not altered by the CEA result. 5. Conclusion PET/CT is far superior to serum CEA for detecting CRC, both preoperatively and during surveillance of disease recurrence. PET/CT detects 2.5 times as many CRCs at initial staging and 1.5 times as many CRC recurrences as serum CEA. Elevated serum CEA levels cannot be relied upon by the clinician to raise the suspicion of disease recurrence. More than a third of all CRC recurrences are CEA negative (even with serial CEA measurements) but detectable by PET/CT. Our results suggest that PET/CT should assume a primary role in the detection of CRC recurrence and that current guidelines recommending only serial CEA measurements and yearly CT abdomen scans need to be revised. References [1] Gold P, Freeman SO. Demonstration of tumor-specific antigens in human colonic carcinomata by immunological tolerance and absorption techniques. J Exp Med 1965;121:439–62.
1097
[2] Wang JY. Prognostic significance of pre- and postoperative serum carcinoembryonic antigen levels in patients with colorectal cancer. Eur Surg Res 2007;39(4): 245–50. [3] Desch CE, Benson AB, Somerfield MR. Colorectal cancer surveillance: 2005 update of an American Society of Clinical Oncology practice guideline. J Clin Oncol 2005; 23:8512–9. [4] Abdel-Nabi H, Doerr RJ, Lamonica DM, Cronin VR, Galantowicz PJ, Carbone GM, et al. Staging of primary colorectal carcinomas with fluorine-18 fluorodeoxyglucose whole body PET: correlation with histopathologic and CT findings. Radiology 1998;206:755–60. [5] Chen LB, Tong JL, Song HZ, Zhu H, Wang YC. 18F-FDG PET/CT in detection of recurrence and metastasis of colorectal cancer. World J Gastroenterol 2007; 13(37):5025–9. [6] Cohade C, Osman M, Leal J, Wahl R. Direct comparison of (18)F-FDG PET and PET/CT in patients with colorectal carcinoma. J Nucl Med 2003;44:1797–803. [7] Moertel CG, Fleming TR, McDonald JS, Haller DG, Laurie JA, Tangen C. An evaluation of the carcinoembryonic antigen (CEA) test for monitoring patients with resected colon cancer. JAMA 1993;270(8):943–7. [8] Ito K, Hibi K, Ando H, Hidemura K, Yamazaki T, Akiyama S, et al. Usefulness of analytical CEA doubling time and half-life time for overlooked synchronous metastases in colorectal carcinoma. Jpn J Clin Oncol 2002;32(2):54–8. [9] Gambhir SS, Czernin J, Schwimmer J, Silverman DHS, Coleman RE, Phelps ME. A tabulated summary of the FDG PET literature. J Nucl Med 2001;42(9):1S–93S (Suppl.). [10] Votrubova J, Belohlavek O, Jaruskova M, Oliverius M, Lohynska R, Trskova K, et al. The role of FDG-PET/CT in the detection of recurrent colorectal cancer. Eur J Nucl Med Mol Imaging 2006;33(7):779–84. [11] Kang S, Seo JH, Bae JH, Lee SW, Yoo J, Ahn BC, et al. F-18 FDG PET/CT can detect recurrence earlier than serum CEA level in patients with colorectal cancer after curative surgical resection. J Nucl Med 2007;48(supplement 2):389P.
Contents lists available at ScienceDirect
Clinical Imaging journal homepage: http://www.clinicalimaging.org
18F-FDG PET/CT superior to serum CEA in detection of colorectal cancer and its recurrence☆ William Makis a,⁎, David Kurzencwyg b, Marc Hickeson b a b
Department of Diagnostic Imaging, Cross Cancer Institute, 11560 University Ave NW Edmonton, AB, T6G 1Z2, Canada Department of Nuclear Medicine, Royal Victoria Hospital, McGill University, 687 Pine Ave West, Floor M2, Montreal, Quebec, H3A 1A1, Canada
a r t i c l e
i n f o
Article history: Received 13 January 2013 Accepted 29 April 2013 Keywords: FDG PET/CT CEA Carcinoembryonic antigen Colorectal cancer
a b s t r a c t The aim of this study was to examine whether positron emission tomography (PET)/computed tomography (CT) can detect more cases of colorectal cancer (CRC) than serum carcinoembryonic antigen (CEA), both at initial staging and during surveillance for recurrence. A retrospective review of 639 CRC patients imaged with PET/CT was performed. PET/CT was superior to serum CEA in detecting CRC, identifying 2.5 times as many CRC at initial staging compared to serum CEA and 1.5 times as many CRC recurrences. The current guideline recommendations of utilizing PET/CT only in the context of a rising serum CEA will miss more than one third of all CRC recurrences. © 2013 Elsevier Inc. All rights reserved.
1. Introduction Since its introduction in 1965, serum carcinoembryonic antigen (CEA) has been an important tumor marker in the management of colorectal cancer (CRC) [1]. CEA does not play a significant role in the staging of CRC due to its low sensitivity [2]; however, serial CEA measurements are recommended by both the 2008 National Comprehensive Cancer Network (NCCN) practice guidelines as well as the 2005 American Society of Clinical Oncology (ASCO) guidelines for the surveillance for CRC recurrence [3]. Positron emission tomography (PET)/computed tomography (CT) is a hybrid imaging modality that takes advantage of the high rate of glycolysis in malignant tumor cells (PET), as well as precise anatomic correlation (CT). PET/CT has not been studied extensively in the initial staging of CRC mainly due to low reported sensitivity in detecting local lymph node involvement [4]. But recently, PET/CT has been reported to have a very high sensitivity and specificity in the detection of CRC recurrence. In a study of 68 patients with CRC recurrence, Chen reported a sensitivity of 94.6% and specificity of 83.3% with PET/CT [5]. Similar results were reported by Cohade [6]. The 2008 NCCN practice guidelines for colorectal cancer suggest performing a PET scan only in the context of an elevated serum CEA.
However, to our knowledge, a direct comparison between PET/CT results and serum CEA has not been done to justify this recommendation. The objective of this paper is to determine if PET/CT can detect more CRC recurrences than elevated serum CEA levels would suggest, and whether a revision of current guidelines for surveillance of CRC recurrence is needed.
2. Materials and methods 2.1. Study design and patient population A retrospective review of 639 patients referred for 18F-FDG PET/CT imaging to the McGill University Health Centre between September 2005 and October 2007 for initial staging of CRC or surveillance of CRC recurrence was performed. Of the 639 patients imaged, 328 had corresponding serum CEA measurements done within 4 months of the PET/CT scan. These were subsequently categorized into two groups: the initial staging group had 189 patients, and the surveillance of CRC recurrence group had 139 patients.
2.2. CEA measurement ☆ This manuscript or any figure or table in it has not been submitted to any publication previously. None of the authors have any financial or other relationships that might lead to a conflict of interest. The manuscript has been read and approved by all the authors, and the requirements for authorship have been met. Each author believes that the manuscript represents honest work. ⁎ Corresponding author. Tel.: +1 780 432 8760; fax: +1 780 432 8411. E-mail address: [email protected] (W. Makis). 0899-7071/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.clinimag.2013.04.004
Between September 2005 and July 24, 2007, serum CEA measurements were done with the CEA-Roche Kit (normal values: 0–4.9). After July 24, 2007, serum CEA measurements were done with the Beckman Coulter Access Immunoassay system (Burlington, Ontario, Canada) (normal values: 0–3.4).
W. Makis et al. / Clinical Imaging 37 (2013) 1094–1097
2.3.
18
F-FDG PET/CT imaging
18
F-FDG PET/CT studies were performed on a hybrid PET/CT scanner (Discovery ST, General Electric Medical Systems, Waukesha, WI, USA), which combines a dedicated, full-ring PET scanner with a 16-slice CT scanner. Patients were required to fast for at least 6 h before the time of their appointment and waited in a quiet dark room the morning prior to their scan. Blood glucose levels were recorded immediately prior to 18F-FDG administration. If the serum glucose level was greater than 11.1 mmol/l (200 mg/dl), then the study was rescheduled. A volume of 400 ml of barium sulfate oral contrast was administered, and 0.22 mCi/kg of 18F-FDG was injected intravenously, up to a maximum dose of 20 mCi. Approximately 60 min following 18 F-FDG injection, CT and PET images were consecutively acquired from the base of the skull to the upper thighs, with additional images of the extremities acquired if needed. For the CT scan portion of the study, the settings were as follows: 140 kVp, 90–110 mA (depending on the body weight), a rotation time of 0.8 s, a table speed of 17 mm per gantry rotation, a pitch of 1.75:1, and a detector row configuration of 16×0.625 mm. For the PET portion of the study, a two-dimensional acquisition was performed, images were acquired using 4–5 min per bed position (depending on the body weight), and five to six bed positions were used (depending on the patient's height). The patient was allowed to breathe normally during the PET and CT acquisitions. Data obtained from the CT acquisition were used for attenuation correction and fusion with PET images. The PET data were reconstructed iteratively using the ordered subset expectation maximization software provided by the manufacturer (21 subsets, 2 iterations). PET attenuation corrected, PET non-attenuation corrected, CT, and PET/CT fusion images of the whole body were displayed in the axial, coronal, and sagittal planes and were reviewed on a dedicated workstation (Xeleris 2.0, GE Healthcare, Waukesha, WI, USA). PET data were also displayed in a rotating maximum intensity projection (MIP) image. 2.4. Interpretation and analysis of PET/CT images All PET/CT images were interpreted using visualization and semiquantitative analysis (SUVmax corrected for body weight) by two nuclear medicine physicians—who have a broad experience in clinical PET/CT and are familiar with normal variants, artifacts, and pitfalls in the interpretation of PET/CT studies—independently. PET/CT results were classified as positive or suspicious for neoplasm versus negative or not suspicious for neoplasm. 3. Results From the 189 patients referred for initial staging of CRC, PET/CT detected approximately 2.5 times (n=182) as many CRCs as CEA (n= 73) (Table 1). All 182 PET/CT-positive cases had histologically proven CRC. Seven patients had both negative PET/CT and CEA, and subsequent workup did not reveal any evidence of CRC in these patients. From the 139 patients referred for surveillance of CRC recurrence, PET/CT detected approximately 1.5 times (n=99) as many CRC recurrences as CEA (n=65). All 99 PET/CT positive cases had radiologic or histologic confirmation of CRC recurrence. Ten patients had a negative PET/CT with a positive CEA, eight of which (80%) did not show Table 1 PET/CT and CEA results for staging scans and surveillance scans of 328 CRC patients Staging scans (n=189)
CEA + CEA −
Surveillance scans (n=139)
PET +
PET −
PET +
PET −
73 109
0 7
65 34
10 30
1095
any evidence of disease recurrence in the subsequent 18 months of clinical and radiological follow-up. One patient had a brain metastasis, which was not detected by PET/CT as it was not in the image acquisition range; however, this finding was known from prior clinical and radiologic evaluation. Another patient had a negative PET/CT with a borderline-elevated CEA (3.6 with normal range of 0–3.4). However, the PET/CT became positive on a subsequent scan done 3 months after the initial negative scan. None of the PET/CT-negative and CEA-positive cases were mucinous adenocarcinomas histologically. 4. Discussion 4.1. PET/CT as initial diagnostic staging modality Serum CEA has a very low sensitivity in the diagnostic workup of CRCs. In a recent study of 425 CRC patients, only 42.6% of all diagnosed CRC patients had an elevated preoperative CEA [2]. In our series, PET/ CT detected 2.5 times as many CRC at initial staging as serum CEA (Fig. 1). These findings suggest that despite the shortcomings of its predecessor (PET), PET/CT's role in the initial staging of CRC warrants closer examination. 4.2. PET/CT in detection of CRC recurrence and metastases Recurrence of CRC occurs in about one third of patients within the first 2 years after colonic resection or rectal amputation and is most frequent in the area contiguous with the surgery. The ASCO 2005 guideline for stage II and III CRC patients recommends measuring serum CEA every 3 months and performing CT scans of the abdomen yearly [3]. However, CEA also has a low sensitivity in the detection of CRC recurrence. Moertel et al. reported a sensitivity of 59% and specificity of 84% for serum CEA in detecting CRC recurrence [7]. Recent attempts have been made to increase the sensitivity of serial CEA measurements by various mathematical manipulations, including calculating a CEA doubling time as well as a CEA half-time, with uncertain results [8]. However, even if CEA detects CRC recurrence, it does not provide any information about the location of the recurrence. CT also has suboptimal sensitivity and specificity in the detection of CRC recurrence, with one meta-analysis reporting a sensitivity of 79% and specificity of 73% [9]. Initial studies done with 18F-FDG PET validated its use in detection of hepatic metastases from CRC, but PET's role in local and regional disease recurrence was less clear. The modality proved to be problematic, as abnormalities could not always be anatomically localized and the rate of false positives was significant. The hybrid PET/CT imaging modality has been reported to significantly increase the sensitivity of PET in the detection of recurrence of CRC. The CT adds excellent anatomical resolution (the increased sensitivity is the result of CT's ability to detect small lung and liver metastases) and physiologic or inflammatory glucose uptake can usually be correctly interpreted on the basis of typical CT findings, which increases specificity. Votrubova et al. showed an increase in sensitivity and specificity from 80% and 69% for PET to 89% and 92% for PET/CT in detecting CRC recurrence in 84 patients [10]. Chen et al. found the sensitivity and specificity of PET/CT to be 94.6% and 83.3% in their 68 patient study [5]. Despite these excellent results, PET/CT is not yet considered an evidence-based tool in the diagnosis of CRC recurrence. 4.3. Positive PET/CT with negative CEA (for CRC recurrence) Our results showed that PET/CT detected 1.5 times as many CRC recurrences as CEA alone. Hence, any protocol or guideline recommendation of performing PET/CT only in the context of a positive CEA will miss a large number (34.3%) of patients who have CRC recurrence, with completely normal CEA levels, even on serial CEA measurements (Fig. 2).
1096
W. Makis et al. / Clinical Imaging 37 (2013) 1094–1097
Fig. 1. Staging PET/CT positive, CEA negative. A 58-year-old man imaged with a staging PET/CT preoperatively for colorectal carcinoma had an FDG-positive rectal mass and a liver metastasis (arrows), both confirmed on biopsy. Serum CEA level done 1 week prior to PET/CT was in the normal range (0.7; normal: 0–4.9). (A) Coronal CT, (B) PET, (C) PET/CT fusion, and (D) MIP images showing FDG-positive rectal mass and liver metastasis.
An argument could be made that PET/CT results should only be compared to CEA results when evaluating tumors that have elevated pretreatment CEA levels to begin with. Of our 39 surveillance PET scans that were PET positive and CEA negative, preoperative CEA levels were not available; however, 21 had serial CEA measurements done posttreatment. Of these, four (19%) became CEA positive with a lag time of 4–12 months following the positive PET/CT result. A similar result was observed by Kang et al., who followed CRC patients, all of whom had positive preoperative serum CEA levels. Following therapy, 69 of their patients had a positive PET/CT scan, and 20% of them had negative CEA levels [11]. PET/CT outperforms serum CEA, even when only looking at CEAproducing colorectal tumors. However, it is important to note that none of the current guidelines make a distinction between CEA- or non-CEA-producing colorectal tumors when determining whether a PET/CT should be done or not. This is another shortcoming of the current guidelines.
4.4. Negative PET/CT with positive CEA Ten of our 139 patients imaged for surveillance of CRC recurrence had a negative PET/CT with a positive CEA, of which 8 did not show any evidence of disease recurrence in the subsequent 18 months of clinical and radiologic follow-up, with CEA levels remaining stable. In the two cases that did show evidence of disease recurrence, one had a pathologically proven brain metastasis, which was not detected by PET/CT as it was not in the image acquisition range. The other case had a borderline-elevated CEA, which became significantly elevated as well as PET/CT positive just 3 months later. None of the PET/CT-negative and CEA-positive cases were mucinous adenocarcinomas histologically. Similar results were reported by Kang et al. from their 171 patients; only three had a negative PET with a positive CEA. None of these three patients had evidence of disease recurrence after further investigation [11]. These results suggest that, in most cases where the CEA is elevated but PET/CT is negative, it is the PET/CT that gives the
Fig. 2. Surveillance PET/CT positive, CEA negative. A 64-year-old man previously diagnosed with colorectal carcinoma, treated surgically 3 years prior, had a follow-up PET/CT after an abdominal CT showed a growing adrenal mass. The PET/CT showed bilateral FDG-positive lung metastases and a right adrenal metastasis (arrows). Serum CEA level done 2 months following PET/CT was in the normal range (3.7; normal: 0–4.9). (A) Coronal CT, (B) PET, (C) PET/CT fusion, and (D) MIP images showing FDG-positive bilateral lung metastases and a right adrenal metastasis.
W. Makis et al. / Clinical Imaging 37 (2013) 1094–1097
correct diagnosis. In our study, serum CEA detected disease that was not detected by PET/CT in only two cases; however, the clinical management in these two cases was not altered by the CEA result. 5. Conclusion PET/CT is far superior to serum CEA for detecting CRC, both preoperatively and during surveillance of disease recurrence. PET/CT detects 2.5 times as many CRCs at initial staging and 1.5 times as many CRC recurrences as serum CEA. Elevated serum CEA levels cannot be relied upon by the clinician to raise the suspicion of disease recurrence. More than a third of all CRC recurrences are CEA negative (even with serial CEA measurements) but detectable by PET/CT. Our results suggest that PET/CT should assume a primary role in the detection of CRC recurrence and that current guidelines recommending only serial CEA measurements and yearly CT abdomen scans need to be revised. References [1] Gold P, Freeman SO. Demonstration of tumor-specific antigens in human colonic carcinomata by immunological tolerance and absorption techniques. J Exp Med 1965;121:439–62.
1097
[2] Wang JY. Prognostic significance of pre- and postoperative serum carcinoembryonic antigen levels in patients with colorectal cancer. Eur Surg Res 2007;39(4): 245–50. [3] Desch CE, Benson AB, Somerfield MR. Colorectal cancer surveillance: 2005 update of an American Society of Clinical Oncology practice guideline. J Clin Oncol 2005; 23:8512–9. [4] Abdel-Nabi H, Doerr RJ, Lamonica DM, Cronin VR, Galantowicz PJ, Carbone GM, et al. Staging of primary colorectal carcinomas with fluorine-18 fluorodeoxyglucose whole body PET: correlation with histopathologic and CT findings. Radiology 1998;206:755–60. [5] Chen LB, Tong JL, Song HZ, Zhu H, Wang YC. 18F-FDG PET/CT in detection of recurrence and metastasis of colorectal cancer. World J Gastroenterol 2007; 13(37):5025–9. [6] Cohade C, Osman M, Leal J, Wahl R. Direct comparison of (18)F-FDG PET and PET/CT in patients with colorectal carcinoma. J Nucl Med 2003;44:1797–803. [7] Moertel CG, Fleming TR, McDonald JS, Haller DG, Laurie JA, Tangen C. An evaluation of the carcinoembryonic antigen (CEA) test for monitoring patients with resected colon cancer. JAMA 1993;270(8):943–7. [8] Ito K, Hibi K, Ando H, Hidemura K, Yamazaki T, Akiyama S, et al. Usefulness of analytical CEA doubling time and half-life time for overlooked synchronous metastases in colorectal carcinoma. Jpn J Clin Oncol 2002;32(2):54–8. [9] Gambhir SS, Czernin J, Schwimmer J, Silverman DHS, Coleman RE, Phelps ME. A tabulated summary of the FDG PET literature. J Nucl Med 2001;42(9):1S–93S (Suppl.). [10] Votrubova J, Belohlavek O, Jaruskova M, Oliverius M, Lohynska R, Trskova K, et al. The role of FDG-PET/CT in the detection of recurrent colorectal cancer. Eur J Nucl Med Mol Imaging 2006;33(7):779–84. [11] Kang S, Seo JH, Bae JH, Lee SW, Yoo J, Ahn BC, et al. F-18 FDG PET/CT can detect recurrence earlier than serum CEA level in patients with colorectal cancer after curative surgical resection. J Nucl Med 2007;48(supplement 2):389P.