A General Cutoff Level Combined With Personalized Dynamic Change of Serum Carcinoembryonic Antigen Can Suggest Timely Use of FDG PET for Early Detection of Recurrent Colorectal Cancer Yu-Yi Huang, MD,* Pei-Ing Lee, MD,* Mei-Ching Liu, MD,† Chien-Chih Chen, MD,‡ Kuo-Cheng Huang, MD,† and Andrew T. Huang, MD† Purpose: FDG PET that has been used is good for diagnosing asymptom- atic colorectal cancer (CRC) recurrence in patients with elevated serum carcinoembryonic antigen (CEA) level. However, there is no reference level of CEA rise that would universally suggest the necessity of a PET study. The purpose of this retrospective study was to identify the high-risk group of CRC recurrence through an examination of the dynamics of the CEA level rise as a recurrence indicator. Patients and Methods: Between July 2002 and May 2010, 112 patients (59 men, 53 women; age, 18–87 years) had FDG PET for suspicious CRC recurrence indicated by elevated CEA level. We reviewed the PET results and the medical records for recurrence verification and calculated the ratio of increase and the velocity of change in CEA levels for risk stratification. Results: The patient-based sensitivity, specificity, and accuracy of PET are 96.6%, 91.3%, and 95.5%, respectively. The probability of recurrence pos- itively correlated with the CEA level rise and the newly diagnosed disease stage. Carcinoembryonic antigen level greater than 13 ng/mL indicated sig- nificantly higher risks of recurrence. In patients with CEA level rise of 13 ng/mL or less, an increase over 3.34 times the individualized baseline also indicated high risks of recurrence. Conclusions: A posttreatment CEA level rise to greater than 13 ng/mL is suggestive of the optimal use of FDG PET, and so is a mild increase below 13 ng/mL at an increase rate over 3.34. Key Words: colorectal cancer, 18 F-FDG PET, carcinoembryonic antigen (Clin Nucl Med 2015;40: e465–e469) E arly detection of asymptomatic recurrence of colorectal cancer (CRC), the third most common cause of cancer death in Taiwan and worldwide, can increase the chance for an effective sec- ond attempt at therapy. 1–3 Studies showed as many as 13% to 48% of patients presented with the nonmetastatic stage II or III disease are likely to develop recurrence despite the effort of curative pri- mary treatment with or without adjuvant therapy. 4–8 Thus adequate posttreatment surveillance is warranted for detection of early recurrence. Carcinoembryonic antigen (CEA) test has long been used to screen for certain types of cancers, especially cancer of the large in- testine (CRC), and according to a large series, CEA was the first in- dication of recurrent disease in 31% of CRC patients with previous surgical resection with curative intent. 9 However, benign elevation of CEA level is not uncommon; it could be related to smoking, gastritis, peptic ulcer disease, diverticulitis, liver disease, chronic obstructive pulmonary disease, diabetes, and any acute or chronic inflammatory state. 10–16 Therefore, further evaluation is necessary, especially in the CRC patients with posttherapy elevated CEA level; the added imaging or laboratory tests can also help identify the site of recurrent disease. The additional testing may include abdominal and pelvic CT, chest CT, colonoscopy, and, in some cases, 18 F-FDG PET scan. FDG PET was proven to have excellent performance in detecting metastatic CRC. 17–19 In a meta-analysis, Maas et al 20 con- cluded that the FDG PET was the most accurate whole-body imag- ing modality for assessing recurrent CRC compared with CT or MRI. Several other studies on the utility of FDG PET for pa- tients with posttherapy CEA level rise seen in follow-up examina- tions also disclosed the superiority of FDG PET over the conventional imaging studies. 21–27 Between July 2002 and May 2010, more than 100 patients had FDG PET examinations on account of rising serum CEA levels, which indicated possible recurrent disease. Although we observed the high sensitivity of FDG PET in detecting sites of recurrent CRC in these patients, we also noted the inconsistency in efficacy of the FDG PET between patients with greater CEA level rise and those with milder elevations. We ventured to assume there is a dif- ference in the patterns of the CEA level elevation, which when iden- tified, may help differentiate a benign from a malignant elevation. Through retrospective assessment of the FDG PET perfor- mance in recurrent CRC detection in patients with elevated post- treatment CEA serum level, we aimed to find predictors that could identify the higher risk groups to ensure the timely use of FDG PET scan during the follow-up examinations. PATIENTS AND METHODS The patient inclusion criteria were as follows: (1) CRC with previous resection treatment of curative intent; (2) CEA level eleva- tion during posttreatment follow-ups; (3) FDG PET surveillance; and (4) histological, cytological results, or follow-up image ob- tained for confirmation of disease nature. The minimum follow-up period was 18 months. PET scan was performed with a PET/CT scanner (GE, Discover LS), and all scans were done 1 hour after the IV injection of 8 to 15 mCi of FDG. Between July 2002 and May 2010, 133 patients were given FDG PET scans on account of suspicious recurrent CRC related to elevated serum CEA level. Results of the PET scans were re- viewed in detail in conjunction with their medical records. Of the 133 patients, 11 with negative PET results were lost to follow-up; 10 patients (9 positive and 1 negative) went on to chemotherapy with no histological or cytological examination done. These 21 patients were not included because there was no way to confirm whether their PET result, positive or negative, was a true or false one. A final total of 112 patients (59 men, 53 women; age, 18–87 years) were included in this study (Table 1). To confirm the presence or Received for publication November 11, 2014; revision accepted May 17, 2015. From the Departments of *Nuclear Medicine, †Medical Oncology, and ‡Surgery, Koo-Foundation, Sun Yat-Sen Cancer Center, Taipei, Taiwan. Conflicts of interest and sources of funding: none declared. Correspondence to: Yu-Yi Huang, MD, Department of Nuclear Medicine, Koo-Foundation, Sun Yat-Sen Cancer Center, 125 Li-Der Rd, Beitou District, Taipei 112, Taiwan. E-mail: yuyi1976@gmail.com, yuyi@kfsyscc.org. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0363-9762/15/4010–e465 DOI: 10.1097/RLU.0000000000000900 ORIGINAL ARTICLE Clinical Nuclear Medicine • Volume 40, Number 10, October 2015 www.nuclearmed.com e465 Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.