- Email: [email protected]
Diagnostic and prognostic values of FDG-PET in patients with non-small cell lung cancer
Clinical Imaging 33 (2009) 90 – 95
Diagnostic and prognostic values of FDG-PET in patients with non-small cell lung cancer Koichiro Abe a,⁎, Shingo Baba a , Koichiro Kaneko a , Takuro Isoda a , Hidetake Yabuuchi a , Masayuki Sasaki b , Shuji Sakai b , Ichiro Yoshino c , Hiroshi Honda a a
Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan b Department of Health Sciences, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan c Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan Received 2 June 2008; accepted 17 June 2008
Abstract Purpose: The aim of this study was to address the efficacy of 2-[F-18]-Fluoro-2-deoxy-D-glucose positron emission tomography (FDGPET) in the staging of non-small cell lung cancer (NSCLC) and in prognostic prediction in patients with NSCLC. Methods: Forty-four patients (26 males, 18 females) were analyzed. Results: Accurate staging was obtained by addition of FDG-PET. Multivariate analysis indicated that the standardized uptake value of the primary tumor was the most significant prognostic factor for disease-free survival (P=.0073). Conclusion: FDG-PET is useful for the diagnosis of NSCLC and for prognostic prediction in patients with NSCLC. © 2009 Elsevier Inc. All rights reserved. Keywords: FDG-PET; Non-small cell lung cancer; Staging; Prognosis; SUV
1. Introduction Lung cancer is one of the most common malignant tumors in the world, and it has been the leading cause of deaths from malignant neoplasms in Japan since 1998 [1]. Staging of lung cancer is considered to be the most important factor for determining optimal treatment and for predicting prognosis [2]. Various imaging techniques, including chest radiography, computed tomography (CT), magnetic resonance imaging (MRI), and others, have been employed for the staging of lung cancer. However, they have limitations because they intrinsically provide only structural information. 2-[F-18]-Fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET), on the other hand, is a functional imaging technique that permits characterization of tumor metabolism. FDG-PET has played an important role in the ⁎ Corresponding author. Department of Clinical Radiology, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan. Tel.: +81 92 642 5695; fax: +81 92 642 5708. E-mail address: [email protected] (K. Abe). 0899-7071/08/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.clinimag.2008.06.032
staging of various malignant tumors, including colorectal cancer, head and neck cancer, breast cancer, and non-small cell lung cancer (NSCLC) [3–6]. The clinical efficacy of FDGPET in the workup of patients with NSCLC has been proven for characterization of lung nodules, evaluation of mediastinal lymph node (LN) staging, and detection of unexpected distant metastases [7]. It is expected that FDG-PET can also provide prognostic information on NSCLC patients, since FDG uptake should reflect the aggressiveness of cancer cells. In this study, we retrospectively reviewed 44 NSCLC patients, all of whom were treated with surgery. The diagnostic performance and prognostic value of FDG-PET were analyzed.
2. Materials and methods 2.1. Patients Forty-four consecutively presenting patients (26 males, 18 females) were retrospectively analyzed in this study. The
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
patients' ages ranged from 36 to 82 years (mean, 64± 11 years). All 44 patients were initially treated with surgery. Patients were excluded if they had received any therapy or surgical diagnostic procedure before FDG-PET. If patients had a previous history of other cancers, they were required to have been disease-free for at least 2 years. Patients whose blood glucose level was more than 150 mg/dl before FDG injection were also excluded. Thirty-six patients had adenocarcinoma, four patients had squamous cell carcinoma, two patients had large cell carcinoma, one patient had adenosquamous cell carcinoma, and one patient had sarcomatoid carcinoma. The pathological stages of all patients were determined after surgery, with the number of patients at each stage determined as follows: Stage IA, 21 patients; Stage IB, 10 patients; Stage IIA, 1 patient; Stage IIB, 4 patients; Stage IIIA, 6 patients; and Stage IIIB, 2 patients. This study was approved by the Committee for the Clinical Application of Cyclotron-Produced Radionuclides at Kyushu University Hospital. 2.2. FDG-PET FDG-PET studies were performed with an ECAT EXACT HR + (Siemens, Knoxville, TN, USA). The intrinsic resolution was 4.6 mm full-width at half-maximum at the center. All patients fasted for at least 4 h before the injection of FDG (185 MBq, 5 mCi). Approximately 60 min after the FDG injection, an emission scan on three-dimensional mode was obtained from head to thigh in nine bed positions (2 min for each bed position). After the emission scan, a transmission scan on two-dimensional mode was obtained using a Ga-68/Ge-68 rod source (2 min for each bed position). Attenuation-corrected images were reconstructed with an ordered subset expectation maximization algorithm (2 iterations with 16 ordered subsets) using the segmented attenuation correction method. 2.3. Staging All 44 patients were examined by chest radiography, bone scintigraphy, contrast-enhanced CT, and FDG-PET before surgery. CT scans were performed at 5- to 7-mm intervals from the lung apices to the lower margins of the liver with intravenous administration of nonionic contrast medium (100 ml of iodine or iohexol, at 1.5 ml/s). Clinical staging in each patient was determined using the findings from several imaging techniques with FDG-PET (PET+) and without FDG-PET (PET− ). The intervals between FDG-PET and CT were 1–21 days (median, 9 days). PET− staging was performed by an experienced boardcertificated radiologist blinded to the FDG-PET findings. The diameter of the primary lung lesion was defined as the maximal length of the tumor on transaxial CT image. For evaluation without FDG-PET imaging, LNs with a short axis
91
larger than 10 mm were considered positive for metastasis on the transaxial CT image. PET image interpretation was visually performed by two experienced nuclear medicine physicians blinded to all clinical data. Standardized uptake value (SUV) was calculated as regional radioactivity concentration divided by the injected amount of radioactivity normalized to body weight. The region of interest (ROI) for SUV calculation was manually drawn as small as possible around a focal increased FDG uptake relative to the background on the transaxial image, and mean SUV was used for the analysis. If no accumulation of FDG was visible on the FDG-PET image, the ROI was determined with reference to the corresponding CT image. 2.4. Follow-up After FDG-PET, all patients were treated surgically. The intervals between PET and surgery were 1–77 days (median, 18 days). Chemotherapy was performed in two patients (4.5%) until the end of the observation period. All patients were followed for at least 230 days after the surgery, unless they were earlier found to have recurrence. Recurrences were diagnosed clinically and pathologically. The median observation period was 875.5 days (91–2352 days). Disease-free survival time was defined as the time between the date of surgery and the date of detection of relapse or the date of the last follow-up visit. 2.5. Statistics Univariate survival analyses were performed with the Kaplan–Meier method, and groups were compared using log-rank and Wilcoxon tests. A Cox proportional-hazards model was used for multivariate analysis to assess the effect of the patients' characteristics and other prognostic factors of significance on the end points. All statistical evaluations were performed using the JMP program package (SAS Institute, Cary, NC, USA). Provability levels of b.05 were considered significant.
3. Results 3.1. Diagnostic performance Clinical cancer staging, as determined with PET+ and PET− imaging, was assessed in comparison with pathological staging defined after surgery (Table 1). Thirty-six patients (81.8%) were accurately diagnosed by conventional imaging without FDG-PET. With the addition of FDG-PET findings, one patient was accurately upstaged (N staging) and three patients were downstaged (N staging in two patients and T staging in one patient). Thus, the correct diagnostic rate for overall cancer staging was improved to 90.9% (40 of 44) on PET+ evaluation. Four of 44 cases (9.1%) were incorrectly
92
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
Table 1 Diagnosis of overall cancer staging
diagnosed even with the addition of FDG-PET findings. N status was underestimated in all these patients. Three of these four patients exhibited adenocarcinoma, and one patient had giant cell carcinoma. Fig. 1 shows a case that was upstaged by the addition of FDG-PET findings. A pretracheal LN was found in a patient with adenocarcinoma. However, the LN was regarded to be negative for metastasis because its shortaxis diameter was not more than 10 mm on the CT image. Thus, the clinical stage of this case was assumed to be T1N0M0 (Stage IA) on PET− evaluation. However, high FDG accumulation was observed in this LN on PET scan. The clinical stage was considered to be T1N2M0 (Stage IIIA) after the addition of the FDG-PET findings. After surgery, pretracheal and right hilar LNs were revealed to be positive for metastasis, and postoperative staging was in agreement with the PET+ staging. Fig. 2 shows a patient with left lung adenocarcinoma. Pleural dissemination was suspected on CT because of the irregularity of the thickened left pleura. The clinical stage without FDG-PET findings was supposed to be T4N0M0 (Stage IIIB). When FDG-PET was performed, no significant accumulation of FDG was found on the left pleura. Pleural
Fig. 1. A case upstaged by the addition of FDG-PET findings. A 63-year-old man with adenocarcinoma in the right lung showed pretracheal LN on CT (A). Although the diameter of the short axis of this LN was 8 mm, intense FDG accumulation (SUV=4.24) was observed on PET scan (B). The TNM staging of this case was considered to be T1N2M0 (Stage IIIA), which was consistent with postsurgical staging.
Fig. 2. A case downstaged by the addition of FDG-PET findings. A 60-yearold man presented with adenocarcinoma in the hilum of the left lung. The left pleura was irregularly thickened on CT and considered as pleural dissemination (A). FDG-PET showed no significant accumulation in the left pleura (B). Pathological examination after the surgery and follow-up studies demonstrated no evidence of malignant lesions on the pleura.
metastasis or dissemination was considered to be negative, and the clinical stage of this case was thought to be T2N0M0 (Stage IIB) on PET+ evaluation. Pathological examination after surgery and follow-up studies demonstrated no evidence of malignant lesions on the pleura—a finding that was also consistent with the PET+ evaluation. 3.2. Disease-free survival At the time of the analysis, recurrence was identified in 10 of 44 patients (22.7%). Two cases were diagnosed with recurrence by the second surgery, and the other eight patients were diagnosed with recurrence by clinical follow-ups including CT, MRI, or FDG-PET. The discriminative cutoff of SUV for the FDG accumulation in the primary tumor was defined in terms of disease-free survival, and a cutoff of 3.5
Fig. 3. Survival plots according to the SUV of the primary tumor. The disease-free survival of patients with low SUVs (≤3.5) was much better than that of patients with high SUVs (N3.5). The difference was statistically significant between both groups (Pb.0001).
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
gave the most significantly discriminative log-rank P value. Fig. 3 shows a Kaplan–Meier plot for disease-free survival. The disease-free survival of patients with low SUVs (≥3.5) was much better than that of patients with high SUVs (N3.5), and the difference between the two groups was statistically significant (Pb.0001). The median survival times were 285.5 and 1026 for patients with high and low SUVs, respectively. In order to address the overall prognostic prediction value of FDG-PET, we next depicted the disease-free survival curves based on cancer staging performed with PET− and PET+ evaluations. Fig. 4 shows that the difference in disease-free survival between NSCLC patients in Stage I and NSCLC patients in Stage II or higher tended to be wider in PET+ staging than in PET− staging. Although the log-rank and Wilcoxon tests showed no statistical significance between these groups in both PET+ and PET− staging, it was suggested that a more accurate clinical staging, in terms of prognostic prediction, was obtained by PET+ evaluation than by PET− evaluation.
93
Table 2 Analyses of prognostic factors for disease-free survival Multivariate analysis
Monovariate analysis P P
Variable
n
Age (≤65 vs. N65 years) pStage (Stage I vs. ≥Stage II) Diameter (≤25 vs. N25 mm) N stage (N0 vs. N1, N2, N3) SUV (≤3.5 vs. N3.5)
21/23 .81 31/13 .0004 25/19 .029 34/10 .0057 30/14 b.0001
Relative risk
.95 1.05 .15 5.37 .38 0.36 .96 1.05 .0073 13.5
3.3. Prognostic factors We next performed separate univariate survival analyses for each of the categorical variables in Table 2. Pathological stage defined after surgery (bStage I vs. ≥Stage II), maximum diameter of the primary tumor (≤25 vs. N25 mm), N status (N0 vs. ≥N1), and SUV (≤3.5 vs. N3.5) were found to be significant prognostic prediction factors in univariate survival analysis. In order to address the interaction and joint effect of these categorized variables on disease-free survival, multivariate survival analysis was also performed using a Cox proportional-hazards model. Table 2 shows that only the SUV of the primary tumor was determined to be a significant prognostic factor for diseasefree survival (P=.0073), but the other variables were not strong factors in multivariate survival analysis. As a consequence, the SUV of the primary tumor was found to be the strongest prognostic factor for disease-free survival in both univariate and multivariate analyses. 4. Discussion
Fig. 4. Disease-free survival curves based on PET− and PET+ staging. Disease-free survival curves based on PET− (A) and PET+ (B) staging were depicted. The difference in disease-free survival between Stages I and II was wider on PET+ evaluation than on PET− evaluation. However, no statistical significance was observed by log-rank and Wilcoxon tests.
CT scan plays a major role in the diagnosis and staging of NSCLC. However, it has an intrinsic limitation because it provides morphologic information only. FDG-PET, on the other hand, provides information about glucose metabolism that potentially permits discrimination between benign and malignant lesions. The clinical efficacy of FDG-PET is now widely accepted, and it provides important information on the diagnosis and staging of lung cancer. The value of FDGPET in the diagnosis and staging of NSCLC has been extensively analyzed. Shim et al. [8] demonstrated that FDGPET/CT is significantly better than stand-alone CT for the accurate diagnosis and staging of lung cancer, especially in the evaluation of LN status. A recent meta-analysis by Birim et al. [9] demonstrated that FDG-PET is more accurate than CT imaging for the evaluation of mediastinal LN metastases in patients with NSCLC. They reported that estimates of the overall sensitivity and specificity of FDG-PET for detecting mediastinal LN metastases were 83% and 92%, respectively, while those of CT scan were 59% and 78%, respectively. In our study, the stages of four patients were diagnosed correctly on PET+ evaluation, but were diagnosed incorrectly on PET− evaluation. In three of these four patients, the
94
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
N status was correctly evaluated by adding the FDG-PET findings (Table 1). This result is in line with the above reports. Our results also showed that cancer staging was incorrectly determined in four patients (9.1%) by both PET+ and PET− evaluations. N status was underestimated in all these patients even with the addition of the FDG-PET findings, indicating that careful management is required in the diagnosis of metastatic LNs. Takamochi et al. [10] studied the cause of false-positive and false-negative FDGPET findings in the evaluation of LN staging in NSCLC patients. They reported the limitation of spatial resolution as the cause of false-negative PET findings in 12 of 14 patients. One patient was correctly diagnosed with negative pleural dissemination by the addition of the FDG-PET findings (Fig. 2). There is no question that CT is superior to FDG-PET in the evaluation of primary tumor extension, since its spatial resolution is much higher than that of FDG-PET. However, additional information about metabolic activity does improve the diagnostic accuracy of suspected lesions. Christensen et al. [11] concluded that coupling the high sensitivity of CT with the high specificity of FDG-PET provided accurate characterization of solitary pulmonary nodules. Birim et al. [9] also reported that the T status of patients was diagnosed more accurately with FDG-PET/CT (86%) than with CT alone (79%). Combining conventional CT-based imaging with FDG-PET findings improved the accuracy of diagnosis. When the disease-free survival curves were depicted based on PET− and PET+ staging, the difference in survival between patients in Stage I and patients in Stage II or higher tended to be wider on PET+ evaluation than on PET− evaluation (Fig. 4). Although no statistical significance was observed on PET+ and PET− evaluations, it was suggested that a more accurate clinical staging was obtained by PET+ evaluation than by PET− evaluation. Recently, Kramer et al. [12] analyzed the prognostic value of PET staging that was obtained from PET images alone without any clinical data. They reported that Cox regression analysis found that PET staging was the most significant prognostic factor for survival in patients with NSCLC. This result supports our data. FDG accumulation in lung cancer is shown to correlate with tumor proliferation and aggressiveness [13,14]. FDGPET is expected to be valuable in prognostic prediction and accurate diagnosis. Our data here indicated that the glucose metabolic status of primary tumors closely correlated with patients' prognoses. Patients with high SUVs showed significantly lower disease-free survival rates than those with low SUVs (Pb.0001) (Fig. 3). Multivariate analysis also demonstrated that SUV was the most significant independent factor for disease-free survival (Table 2). To date, many researchers have reported the correlation between FDG uptake in the primary tumor and patients' prognosis [15–22]. Ahuja et al. [15] retrospectively analyzed 155 patients with NSCLC and demonstrated that an SUV of N10 correlated with poorer survival. Eschmann et al. [16] indicated that a cutoff SUV of 12.0 was the most discriminative value for
prognosis in 159 patients with NSCLC. The SUV cutoff of our data is relatively lower than those of published results. This can be explained by the fact that we examined patients with mostly early-stage lung cancer who were treated with surgery, while the previous studies, except for two, included patients with diseases at advanced stages who were treated by chemotherapy and/or radiotherapy. There are two reports that have analyzed early-stage NSCLC [17,18]. The authors of these reports demonstrated that the cutoff SUVs for disease-free survival were 5 and 3.3, respectively—results similar to ours. Another reason for the difference might be that we used mean SUV. Although we drew the ROI on the small area of focal FDG uptake, and not on the whole area in the primary tumor, the influence of partial-volume effect could not be avoided. The SUV as a semiquantitative parameter of glucose uptake can be affected by various factors such as patient size, accurate injection dose, plasma glucose and insulin levels, the interval between FDG injection and image acquisition, partial-volume effects, and so on. A recent report divided 498 lung cancer patients into five subgroups according to the SUVs of their primary tumors and analyzed the relationship between these SUV subgroups and survival rates [19]. They demonstrated a clear association between high FDG accumulation in the tumor and poor prognosis without choosing any cutoff SUV. Because of the differences in the patient populations, the use of different sets of protocols, and many other factors mentioned above, an arbitrary cutoff value cannot be defined. Instead, several factors and conditions for each PET facility should be considered for the determination of a proper SUV cutoff to predict a patient's prognosis. The principal limitation of our study was the small number of patients. Because of this small number, patients who received palliative surgery were also included, in addition to those who received curative treatment. This limitation could result in a large selection bias in this study. Further accumulation of data is needed. In conclusion, the FDG-PET findings provided a diagnostic accuracy of cancer staging higher than those of conventional imaging methods. In addition, the SUVs of primary tumors closely correlated with disease-free survival in patients with NSCLC. FDG-PET proved to be a powerful tool not only for diagnosis but also for prognostic prediction. References [1] Annual Statistical Report of National Health Conditions. Health and welfare statistics association. Tokyo, Japan: Japanese Ministry of Health and Welfare. p. 48. [2] Brundage MD, Davies D, Mackillop WJ. Prognostic factors in nonsmall cell lung cancer: a decade of progress. Chest 2002;122:1037–57. [3] Strauss LG, Clorius JH, Schlag P, Lehner B, Kimmig B, Engenhart R, et al. Recurrence of colorectal tumors: PET evaluation. Radiology 1989;170:329–32. [4] Lapela M, Grenman R, Kurki T, Joensuu H, Leskinen S, Lindholm P, et al. Head and neck cancer: detection of recurrence with PET and 2-[18F]-fluoro-2-deoxy-D-glucose. Radiology 1995;197:135–9.
K. Abe et al. / Clinical Imaging 33 (2009) 90–95 [5] Crippa F, Agresti R, Seregni E, Greco M, Pascali C, Bogni A, et al. Prospective evaluation of fluorine-18-FDG PET in presurgical staging of the axilla in breast cancer. J Nucl Med 1998;39:4–8. [6] Silvestri GA, Tanoue LT, Margolis ML, Barker J, Detterbeck F, American College of Chest Physicians. The noninvasive staging of non-small cell lung cancer: the guidelines. Chest 2003;123(Suppl 1):147S–56S. [7] Baum RP, Hellwig D, Mezzetti M. Position of nuclear medicine modalities in the diagnostic workup of cancer patients: lung cancer. Q J Nucl Med Mol Imaging 2004;48:119–42. [8] Shim SS, Lee KS, Kim BT, Chung MJ, Lee EJ, Han J, et al. Non-small cell lung cancer: prospective comparison of integrated FDG PET/CT and CT alone for preoperative staging. Radiology 2005;236:1011–9. [9] Birim O, Kappetein AP, Stijnen T, Bogers AJ. Meta-analysis of positron emission tomographic and computed tomographic imaging in detecting mediastinal lymph node metastases in nonsmall cell lung cancer. Ann Thorac Surg 2005;79:375–82. [10] Takamochi K, Yoshida J, Murakami K, Niho S, Ishii G, Nishimura M, et al. Pitfalls in lymph node staging with positron emission tomography in non-small cell lung cancer patients. Lung Cancer 2005;47:235–42. [11] Christensen JA, Nathan MA, Mullan BP, Hartman TE, Swensen SJ, Lowe VJ. Characterization of the solitary pulmonary nodule: 18F-FDG PET versus nodule-enhancement CT. Am J Roentgenol 2006;187:1361–7. [12] Kramer H, Post WJ, Pruim J, Groen HJ. The prognostic value of positron emission tomography in non-small cell lung cancer: analysis of 266 cases. Lung Cancer 2006;52:213–7. [13] Duhaylongsod FG, Lowe VJ, Patz EF, Vaughn AL, Coleman RE, Wolfe WG. Lung tumor growth correlates with glucose metabolism measured by fluoride-18 fluorodeoxyglucose positron emission tomography. Ann Thorac Surg 1995;60:1348–52. [14] Vesselle H, Schmidt RA, Pugsley JM, Li M, Kohlmyer SG, Vallires E, et al. Lung cancer proliferation correlates with [F-18]fluorodeoxyglucose uptake by positron emission tomography. Clin Cancer Res 2000; 6:3837–44.
95
[15] Ahuja V, Coleman RE, Herndon J, Patz EF. The prognostic significance of fluorodeoxyglucose positron emission tomography imaging for patients with nonsmall cell lung carcinoma. Cancer 1998; 83:918–24. [16] Eschmann SM, Friedel G, Paulsen F, Reimold M, Hehr T, Budach W, et al. Is standardised (18)F-FDG uptake value an outcome predictor in patients with Stage III non-small cell lung cancer? Eur J Nucl Med Mol Imaging 2006;33:263–9. [17] Higashi K, Ueda Y, Arisaka Y, Sakuma T, Nambu Y, Oguchi M, et al. 18 F-FDG uptake as a biologic prognostic factor for recurrence in patients with surgically resected non-small cell lung cancer. J Nucl Med 2002;43:39–45. [18] Ohtsuka T, Nomori H, Watanabe K, Kaji M, Naruke T, Suemasu K, et al. Prognostic significance of [18F]fluorodeoxyglucose uptake on positron emission tomography in patients with pathologic Stage I lung adenocarcinoma. Cancer 2006;107:2468–73. [19] Davies A, Tan C, Paschalides C, Barrington SF, O'Doherty M, Utley M, et al. FDG-PET maximum standardised uptake value is associated with variation in survival: analysis of 498 lung cancer patients. Lung Cancer 2007;55:75–8. [20] Downey RJ, Akhurst T, Gonen M, Vincent A, Bains MS, Larson S, et al. Preoperative F-18 fluorodeoxyglucose–positron emission tomography maximal standardized uptake value predicts survival after lung cancer resection. J Clin Oncol 2004;22:3255–60. [21] Vansteenkiste JF, Stroobants SG, Dupont PJ, De Leyn PR, Verbeken EK, Deneffe GJ, et al. Prognostic importance of the standardized uptake value on 18F-fluoro-2-deoxy-glucose-positron emission tomography scan in non-small-cell lung cancer: an analysis of 125 cases. Leuven Lung Cancer Group. J Clin Oncol 1999;17:3201–6. [22] Sasaki R, Komaki R, Macapinlac H, Erasmus J, Allen P, Forster K, et al. [18F]Fluorodeoxyglucose uptake by positron emission tomography predicts outcome of non-small-cell lung cancer. J Clin Oncol 2005;23:1136–43.
Diagnostic and prognostic values of FDG-PET in patients with non-small cell lung cancer Koichiro Abe a,⁎, Shingo Baba a , Koichiro Kaneko a , Takuro Isoda a , Hidetake Yabuuchi a , Masayuki Sasaki b , Shuji Sakai b , Ichiro Yoshino c , Hiroshi Honda a a
Department of Clinical Radiology, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan b Department of Health Sciences, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan c Department of Surgery and Science, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan Received 2 June 2008; accepted 17 June 2008
Abstract Purpose: The aim of this study was to address the efficacy of 2-[F-18]-Fluoro-2-deoxy-D-glucose positron emission tomography (FDGPET) in the staging of non-small cell lung cancer (NSCLC) and in prognostic prediction in patients with NSCLC. Methods: Forty-four patients (26 males, 18 females) were analyzed. Results: Accurate staging was obtained by addition of FDG-PET. Multivariate analysis indicated that the standardized uptake value of the primary tumor was the most significant prognostic factor for disease-free survival (P=.0073). Conclusion: FDG-PET is useful for the diagnosis of NSCLC and for prognostic prediction in patients with NSCLC. © 2009 Elsevier Inc. All rights reserved. Keywords: FDG-PET; Non-small cell lung cancer; Staging; Prognosis; SUV
1. Introduction Lung cancer is one of the most common malignant tumors in the world, and it has been the leading cause of deaths from malignant neoplasms in Japan since 1998 [1]. Staging of lung cancer is considered to be the most important factor for determining optimal treatment and for predicting prognosis [2]. Various imaging techniques, including chest radiography, computed tomography (CT), magnetic resonance imaging (MRI), and others, have been employed for the staging of lung cancer. However, they have limitations because they intrinsically provide only structural information. 2-[F-18]-Fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET), on the other hand, is a functional imaging technique that permits characterization of tumor metabolism. FDG-PET has played an important role in the ⁎ Corresponding author. Department of Clinical Radiology, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan. Tel.: +81 92 642 5695; fax: +81 92 642 5708. E-mail address: [email protected] (K. Abe). 0899-7071/08/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.clinimag.2008.06.032
staging of various malignant tumors, including colorectal cancer, head and neck cancer, breast cancer, and non-small cell lung cancer (NSCLC) [3–6]. The clinical efficacy of FDGPET in the workup of patients with NSCLC has been proven for characterization of lung nodules, evaluation of mediastinal lymph node (LN) staging, and detection of unexpected distant metastases [7]. It is expected that FDG-PET can also provide prognostic information on NSCLC patients, since FDG uptake should reflect the aggressiveness of cancer cells. In this study, we retrospectively reviewed 44 NSCLC patients, all of whom were treated with surgery. The diagnostic performance and prognostic value of FDG-PET were analyzed.
2. Materials and methods 2.1. Patients Forty-four consecutively presenting patients (26 males, 18 females) were retrospectively analyzed in this study. The
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
patients' ages ranged from 36 to 82 years (mean, 64± 11 years). All 44 patients were initially treated with surgery. Patients were excluded if they had received any therapy or surgical diagnostic procedure before FDG-PET. If patients had a previous history of other cancers, they were required to have been disease-free for at least 2 years. Patients whose blood glucose level was more than 150 mg/dl before FDG injection were also excluded. Thirty-six patients had adenocarcinoma, four patients had squamous cell carcinoma, two patients had large cell carcinoma, one patient had adenosquamous cell carcinoma, and one patient had sarcomatoid carcinoma. The pathological stages of all patients were determined after surgery, with the number of patients at each stage determined as follows: Stage IA, 21 patients; Stage IB, 10 patients; Stage IIA, 1 patient; Stage IIB, 4 patients; Stage IIIA, 6 patients; and Stage IIIB, 2 patients. This study was approved by the Committee for the Clinical Application of Cyclotron-Produced Radionuclides at Kyushu University Hospital. 2.2. FDG-PET FDG-PET studies were performed with an ECAT EXACT HR + (Siemens, Knoxville, TN, USA). The intrinsic resolution was 4.6 mm full-width at half-maximum at the center. All patients fasted for at least 4 h before the injection of FDG (185 MBq, 5 mCi). Approximately 60 min after the FDG injection, an emission scan on three-dimensional mode was obtained from head to thigh in nine bed positions (2 min for each bed position). After the emission scan, a transmission scan on two-dimensional mode was obtained using a Ga-68/Ge-68 rod source (2 min for each bed position). Attenuation-corrected images were reconstructed with an ordered subset expectation maximization algorithm (2 iterations with 16 ordered subsets) using the segmented attenuation correction method. 2.3. Staging All 44 patients were examined by chest radiography, bone scintigraphy, contrast-enhanced CT, and FDG-PET before surgery. CT scans were performed at 5- to 7-mm intervals from the lung apices to the lower margins of the liver with intravenous administration of nonionic contrast medium (100 ml of iodine or iohexol, at 1.5 ml/s). Clinical staging in each patient was determined using the findings from several imaging techniques with FDG-PET (PET+) and without FDG-PET (PET− ). The intervals between FDG-PET and CT were 1–21 days (median, 9 days). PET− staging was performed by an experienced boardcertificated radiologist blinded to the FDG-PET findings. The diameter of the primary lung lesion was defined as the maximal length of the tumor on transaxial CT image. For evaluation without FDG-PET imaging, LNs with a short axis
91
larger than 10 mm were considered positive for metastasis on the transaxial CT image. PET image interpretation was visually performed by two experienced nuclear medicine physicians blinded to all clinical data. Standardized uptake value (SUV) was calculated as regional radioactivity concentration divided by the injected amount of radioactivity normalized to body weight. The region of interest (ROI) for SUV calculation was manually drawn as small as possible around a focal increased FDG uptake relative to the background on the transaxial image, and mean SUV was used for the analysis. If no accumulation of FDG was visible on the FDG-PET image, the ROI was determined with reference to the corresponding CT image. 2.4. Follow-up After FDG-PET, all patients were treated surgically. The intervals between PET and surgery were 1–77 days (median, 18 days). Chemotherapy was performed in two patients (4.5%) until the end of the observation period. All patients were followed for at least 230 days after the surgery, unless they were earlier found to have recurrence. Recurrences were diagnosed clinically and pathologically. The median observation period was 875.5 days (91–2352 days). Disease-free survival time was defined as the time between the date of surgery and the date of detection of relapse or the date of the last follow-up visit. 2.5. Statistics Univariate survival analyses were performed with the Kaplan–Meier method, and groups were compared using log-rank and Wilcoxon tests. A Cox proportional-hazards model was used for multivariate analysis to assess the effect of the patients' characteristics and other prognostic factors of significance on the end points. All statistical evaluations were performed using the JMP program package (SAS Institute, Cary, NC, USA). Provability levels of b.05 were considered significant.
3. Results 3.1. Diagnostic performance Clinical cancer staging, as determined with PET+ and PET− imaging, was assessed in comparison with pathological staging defined after surgery (Table 1). Thirty-six patients (81.8%) were accurately diagnosed by conventional imaging without FDG-PET. With the addition of FDG-PET findings, one patient was accurately upstaged (N staging) and three patients were downstaged (N staging in two patients and T staging in one patient). Thus, the correct diagnostic rate for overall cancer staging was improved to 90.9% (40 of 44) on PET+ evaluation. Four of 44 cases (9.1%) were incorrectly
92
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
Table 1 Diagnosis of overall cancer staging
diagnosed even with the addition of FDG-PET findings. N status was underestimated in all these patients. Three of these four patients exhibited adenocarcinoma, and one patient had giant cell carcinoma. Fig. 1 shows a case that was upstaged by the addition of FDG-PET findings. A pretracheal LN was found in a patient with adenocarcinoma. However, the LN was regarded to be negative for metastasis because its shortaxis diameter was not more than 10 mm on the CT image. Thus, the clinical stage of this case was assumed to be T1N0M0 (Stage IA) on PET− evaluation. However, high FDG accumulation was observed in this LN on PET scan. The clinical stage was considered to be T1N2M0 (Stage IIIA) after the addition of the FDG-PET findings. After surgery, pretracheal and right hilar LNs were revealed to be positive for metastasis, and postoperative staging was in agreement with the PET+ staging. Fig. 2 shows a patient with left lung adenocarcinoma. Pleural dissemination was suspected on CT because of the irregularity of the thickened left pleura. The clinical stage without FDG-PET findings was supposed to be T4N0M0 (Stage IIIB). When FDG-PET was performed, no significant accumulation of FDG was found on the left pleura. Pleural
Fig. 1. A case upstaged by the addition of FDG-PET findings. A 63-year-old man with adenocarcinoma in the right lung showed pretracheal LN on CT (A). Although the diameter of the short axis of this LN was 8 mm, intense FDG accumulation (SUV=4.24) was observed on PET scan (B). The TNM staging of this case was considered to be T1N2M0 (Stage IIIA), which was consistent with postsurgical staging.
Fig. 2. A case downstaged by the addition of FDG-PET findings. A 60-yearold man presented with adenocarcinoma in the hilum of the left lung. The left pleura was irregularly thickened on CT and considered as pleural dissemination (A). FDG-PET showed no significant accumulation in the left pleura (B). Pathological examination after the surgery and follow-up studies demonstrated no evidence of malignant lesions on the pleura.
metastasis or dissemination was considered to be negative, and the clinical stage of this case was thought to be T2N0M0 (Stage IIB) on PET+ evaluation. Pathological examination after surgery and follow-up studies demonstrated no evidence of malignant lesions on the pleura—a finding that was also consistent with the PET+ evaluation. 3.2. Disease-free survival At the time of the analysis, recurrence was identified in 10 of 44 patients (22.7%). Two cases were diagnosed with recurrence by the second surgery, and the other eight patients were diagnosed with recurrence by clinical follow-ups including CT, MRI, or FDG-PET. The discriminative cutoff of SUV for the FDG accumulation in the primary tumor was defined in terms of disease-free survival, and a cutoff of 3.5
Fig. 3. Survival plots according to the SUV of the primary tumor. The disease-free survival of patients with low SUVs (≤3.5) was much better than that of patients with high SUVs (N3.5). The difference was statistically significant between both groups (Pb.0001).
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
gave the most significantly discriminative log-rank P value. Fig. 3 shows a Kaplan–Meier plot for disease-free survival. The disease-free survival of patients with low SUVs (≥3.5) was much better than that of patients with high SUVs (N3.5), and the difference between the two groups was statistically significant (Pb.0001). The median survival times were 285.5 and 1026 for patients with high and low SUVs, respectively. In order to address the overall prognostic prediction value of FDG-PET, we next depicted the disease-free survival curves based on cancer staging performed with PET− and PET+ evaluations. Fig. 4 shows that the difference in disease-free survival between NSCLC patients in Stage I and NSCLC patients in Stage II or higher tended to be wider in PET+ staging than in PET− staging. Although the log-rank and Wilcoxon tests showed no statistical significance between these groups in both PET+ and PET− staging, it was suggested that a more accurate clinical staging, in terms of prognostic prediction, was obtained by PET+ evaluation than by PET− evaluation.
93
Table 2 Analyses of prognostic factors for disease-free survival Multivariate analysis
Monovariate analysis P P
Variable
n
Age (≤65 vs. N65 years) pStage (Stage I vs. ≥Stage II) Diameter (≤25 vs. N25 mm) N stage (N0 vs. N1, N2, N3) SUV (≤3.5 vs. N3.5)
21/23 .81 31/13 .0004 25/19 .029 34/10 .0057 30/14 b.0001
Relative risk
.95 1.05 .15 5.37 .38 0.36 .96 1.05 .0073 13.5
3.3. Prognostic factors We next performed separate univariate survival analyses for each of the categorical variables in Table 2. Pathological stage defined after surgery (bStage I vs. ≥Stage II), maximum diameter of the primary tumor (≤25 vs. N25 mm), N status (N0 vs. ≥N1), and SUV (≤3.5 vs. N3.5) were found to be significant prognostic prediction factors in univariate survival analysis. In order to address the interaction and joint effect of these categorized variables on disease-free survival, multivariate survival analysis was also performed using a Cox proportional-hazards model. Table 2 shows that only the SUV of the primary tumor was determined to be a significant prognostic factor for diseasefree survival (P=.0073), but the other variables were not strong factors in multivariate survival analysis. As a consequence, the SUV of the primary tumor was found to be the strongest prognostic factor for disease-free survival in both univariate and multivariate analyses. 4. Discussion
Fig. 4. Disease-free survival curves based on PET− and PET+ staging. Disease-free survival curves based on PET− (A) and PET+ (B) staging were depicted. The difference in disease-free survival between Stages I and II was wider on PET+ evaluation than on PET− evaluation. However, no statistical significance was observed by log-rank and Wilcoxon tests.
CT scan plays a major role in the diagnosis and staging of NSCLC. However, it has an intrinsic limitation because it provides morphologic information only. FDG-PET, on the other hand, provides information about glucose metabolism that potentially permits discrimination between benign and malignant lesions. The clinical efficacy of FDG-PET is now widely accepted, and it provides important information on the diagnosis and staging of lung cancer. The value of FDGPET in the diagnosis and staging of NSCLC has been extensively analyzed. Shim et al. [8] demonstrated that FDGPET/CT is significantly better than stand-alone CT for the accurate diagnosis and staging of lung cancer, especially in the evaluation of LN status. A recent meta-analysis by Birim et al. [9] demonstrated that FDG-PET is more accurate than CT imaging for the evaluation of mediastinal LN metastases in patients with NSCLC. They reported that estimates of the overall sensitivity and specificity of FDG-PET for detecting mediastinal LN metastases were 83% and 92%, respectively, while those of CT scan were 59% and 78%, respectively. In our study, the stages of four patients were diagnosed correctly on PET+ evaluation, but were diagnosed incorrectly on PET− evaluation. In three of these four patients, the
94
K. Abe et al. / Clinical Imaging 33 (2009) 90–95
N status was correctly evaluated by adding the FDG-PET findings (Table 1). This result is in line with the above reports. Our results also showed that cancer staging was incorrectly determined in four patients (9.1%) by both PET+ and PET− evaluations. N status was underestimated in all these patients even with the addition of the FDG-PET findings, indicating that careful management is required in the diagnosis of metastatic LNs. Takamochi et al. [10] studied the cause of false-positive and false-negative FDGPET findings in the evaluation of LN staging in NSCLC patients. They reported the limitation of spatial resolution as the cause of false-negative PET findings in 12 of 14 patients. One patient was correctly diagnosed with negative pleural dissemination by the addition of the FDG-PET findings (Fig. 2). There is no question that CT is superior to FDG-PET in the evaluation of primary tumor extension, since its spatial resolution is much higher than that of FDG-PET. However, additional information about metabolic activity does improve the diagnostic accuracy of suspected lesions. Christensen et al. [11] concluded that coupling the high sensitivity of CT with the high specificity of FDG-PET provided accurate characterization of solitary pulmonary nodules. Birim et al. [9] also reported that the T status of patients was diagnosed more accurately with FDG-PET/CT (86%) than with CT alone (79%). Combining conventional CT-based imaging with FDG-PET findings improved the accuracy of diagnosis. When the disease-free survival curves were depicted based on PET− and PET+ staging, the difference in survival between patients in Stage I and patients in Stage II or higher tended to be wider on PET+ evaluation than on PET− evaluation (Fig. 4). Although no statistical significance was observed on PET+ and PET− evaluations, it was suggested that a more accurate clinical staging was obtained by PET+ evaluation than by PET− evaluation. Recently, Kramer et al. [12] analyzed the prognostic value of PET staging that was obtained from PET images alone without any clinical data. They reported that Cox regression analysis found that PET staging was the most significant prognostic factor for survival in patients with NSCLC. This result supports our data. FDG accumulation in lung cancer is shown to correlate with tumor proliferation and aggressiveness [13,14]. FDGPET is expected to be valuable in prognostic prediction and accurate diagnosis. Our data here indicated that the glucose metabolic status of primary tumors closely correlated with patients' prognoses. Patients with high SUVs showed significantly lower disease-free survival rates than those with low SUVs (Pb.0001) (Fig. 3). Multivariate analysis also demonstrated that SUV was the most significant independent factor for disease-free survival (Table 2). To date, many researchers have reported the correlation between FDG uptake in the primary tumor and patients' prognosis [15–22]. Ahuja et al. [15] retrospectively analyzed 155 patients with NSCLC and demonstrated that an SUV of N10 correlated with poorer survival. Eschmann et al. [16] indicated that a cutoff SUV of 12.0 was the most discriminative value for
prognosis in 159 patients with NSCLC. The SUV cutoff of our data is relatively lower than those of published results. This can be explained by the fact that we examined patients with mostly early-stage lung cancer who were treated with surgery, while the previous studies, except for two, included patients with diseases at advanced stages who were treated by chemotherapy and/or radiotherapy. There are two reports that have analyzed early-stage NSCLC [17,18]. The authors of these reports demonstrated that the cutoff SUVs for disease-free survival were 5 and 3.3, respectively—results similar to ours. Another reason for the difference might be that we used mean SUV. Although we drew the ROI on the small area of focal FDG uptake, and not on the whole area in the primary tumor, the influence of partial-volume effect could not be avoided. The SUV as a semiquantitative parameter of glucose uptake can be affected by various factors such as patient size, accurate injection dose, plasma glucose and insulin levels, the interval between FDG injection and image acquisition, partial-volume effects, and so on. A recent report divided 498 lung cancer patients into five subgroups according to the SUVs of their primary tumors and analyzed the relationship between these SUV subgroups and survival rates [19]. They demonstrated a clear association between high FDG accumulation in the tumor and poor prognosis without choosing any cutoff SUV. Because of the differences in the patient populations, the use of different sets of protocols, and many other factors mentioned above, an arbitrary cutoff value cannot be defined. Instead, several factors and conditions for each PET facility should be considered for the determination of a proper SUV cutoff to predict a patient's prognosis. The principal limitation of our study was the small number of patients. Because of this small number, patients who received palliative surgery were also included, in addition to those who received curative treatment. This limitation could result in a large selection bias in this study. Further accumulation of data is needed. In conclusion, the FDG-PET findings provided a diagnostic accuracy of cancer staging higher than those of conventional imaging methods. In addition, the SUVs of primary tumors closely correlated with disease-free survival in patients with NSCLC. FDG-PET proved to be a powerful tool not only for diagnosis but also for prognostic prediction. References [1] Annual Statistical Report of National Health Conditions. Health and welfare statistics association. Tokyo, Japan: Japanese Ministry of Health and Welfare. p. 48. [2] Brundage MD, Davies D, Mackillop WJ. Prognostic factors in nonsmall cell lung cancer: a decade of progress. Chest 2002;122:1037–57. [3] Strauss LG, Clorius JH, Schlag P, Lehner B, Kimmig B, Engenhart R, et al. Recurrence of colorectal tumors: PET evaluation. Radiology 1989;170:329–32. [4] Lapela M, Grenman R, Kurki T, Joensuu H, Leskinen S, Lindholm P, et al. Head and neck cancer: detection of recurrence with PET and 2-[18F]-fluoro-2-deoxy-D-glucose. Radiology 1995;197:135–9.
K. Abe et al. / Clinical Imaging 33 (2009) 90–95 [5] Crippa F, Agresti R, Seregni E, Greco M, Pascali C, Bogni A, et al. Prospective evaluation of fluorine-18-FDG PET in presurgical staging of the axilla in breast cancer. J Nucl Med 1998;39:4–8. [6] Silvestri GA, Tanoue LT, Margolis ML, Barker J, Detterbeck F, American College of Chest Physicians. The noninvasive staging of non-small cell lung cancer: the guidelines. Chest 2003;123(Suppl 1):147S–56S. [7] Baum RP, Hellwig D, Mezzetti M. Position of nuclear medicine modalities in the diagnostic workup of cancer patients: lung cancer. Q J Nucl Med Mol Imaging 2004;48:119–42. [8] Shim SS, Lee KS, Kim BT, Chung MJ, Lee EJ, Han J, et al. Non-small cell lung cancer: prospective comparison of integrated FDG PET/CT and CT alone for preoperative staging. Radiology 2005;236:1011–9. [9] Birim O, Kappetein AP, Stijnen T, Bogers AJ. Meta-analysis of positron emission tomographic and computed tomographic imaging in detecting mediastinal lymph node metastases in nonsmall cell lung cancer. Ann Thorac Surg 2005;79:375–82. [10] Takamochi K, Yoshida J, Murakami K, Niho S, Ishii G, Nishimura M, et al. Pitfalls in lymph node staging with positron emission tomography in non-small cell lung cancer patients. Lung Cancer 2005;47:235–42. [11] Christensen JA, Nathan MA, Mullan BP, Hartman TE, Swensen SJ, Lowe VJ. Characterization of the solitary pulmonary nodule: 18F-FDG PET versus nodule-enhancement CT. Am J Roentgenol 2006;187:1361–7. [12] Kramer H, Post WJ, Pruim J, Groen HJ. The prognostic value of positron emission tomography in non-small cell lung cancer: analysis of 266 cases. Lung Cancer 2006;52:213–7. [13] Duhaylongsod FG, Lowe VJ, Patz EF, Vaughn AL, Coleman RE, Wolfe WG. Lung tumor growth correlates with glucose metabolism measured by fluoride-18 fluorodeoxyglucose positron emission tomography. Ann Thorac Surg 1995;60:1348–52. [14] Vesselle H, Schmidt RA, Pugsley JM, Li M, Kohlmyer SG, Vallires E, et al. Lung cancer proliferation correlates with [F-18]fluorodeoxyglucose uptake by positron emission tomography. Clin Cancer Res 2000; 6:3837–44.
95
[15] Ahuja V, Coleman RE, Herndon J, Patz EF. The prognostic significance of fluorodeoxyglucose positron emission tomography imaging for patients with nonsmall cell lung carcinoma. Cancer 1998; 83:918–24. [16] Eschmann SM, Friedel G, Paulsen F, Reimold M, Hehr T, Budach W, et al. Is standardised (18)F-FDG uptake value an outcome predictor in patients with Stage III non-small cell lung cancer? Eur J Nucl Med Mol Imaging 2006;33:263–9. [17] Higashi K, Ueda Y, Arisaka Y, Sakuma T, Nambu Y, Oguchi M, et al. 18 F-FDG uptake as a biologic prognostic factor for recurrence in patients with surgically resected non-small cell lung cancer. J Nucl Med 2002;43:39–45. [18] Ohtsuka T, Nomori H, Watanabe K, Kaji M, Naruke T, Suemasu K, et al. Prognostic significance of [18F]fluorodeoxyglucose uptake on positron emission tomography in patients with pathologic Stage I lung adenocarcinoma. Cancer 2006;107:2468–73. [19] Davies A, Tan C, Paschalides C, Barrington SF, O'Doherty M, Utley M, et al. FDG-PET maximum standardised uptake value is associated with variation in survival: analysis of 498 lung cancer patients. Lung Cancer 2007;55:75–8. [20] Downey RJ, Akhurst T, Gonen M, Vincent A, Bains MS, Larson S, et al. Preoperative F-18 fluorodeoxyglucose–positron emission tomography maximal standardized uptake value predicts survival after lung cancer resection. J Clin Oncol 2004;22:3255–60. [21] Vansteenkiste JF, Stroobants SG, Dupont PJ, De Leyn PR, Verbeken EK, Deneffe GJ, et al. Prognostic importance of the standardized uptake value on 18F-fluoro-2-deoxy-glucose-positron emission tomography scan in non-small-cell lung cancer: an analysis of 125 cases. Leuven Lung Cancer Group. J Clin Oncol 1999;17:3201–6. [22] Sasaki R, Komaki R, Macapinlac H, Erasmus J, Allen P, Forster K, et al. [18F]Fluorodeoxyglucose uptake by positron emission tomography predicts outcome of non-small-cell lung cancer. J Clin Oncol 2005;23:1136–43.