Recent advances in positron emission tomography (PET)

Transcription

1 Evaluation of Semiquantitative Assessments of Fluorodeoxyglucose Uptake on Positron Emission Tomography Scans for the Diagnosis of Pulmonary Malignancies 1 to 3 cm in Size Yasuomi Ohba, MD, Hiroaki Nomori, MD, PhD, Hidekatsu Shibata, MD, Hironori Kobayashi, MD, PhD, Takeshi Mori, MD, PhD, Shinya Shiraishi, MD, PhD, and Rumi Nakashima, MD, PhD Departments of Thoracic Surgery and Diagnostic Radiology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto; Japanese Red Cross Kumamoto Health Care Center, Kumamoto; and Division of General Thoracic Surgery, Department of Surgery, School of Medicine, Keio University, Tokyo, Japan Background. To determine the optimal method of evaluating fluorodeoxyglucose (FDG) uptake on positron emission tomography (PET) for the diagnosis of pulmonary malignancies, the sensitivity and specificity of visual assessment and the several semiquantitative analyses were compared. Methods. Positron emission tomography data were analyzed for 130 pulmonary nodules from 1 to 3 cm in size (101 malignant and 29 benign nodules). The FDG uptake was measured by maximum standard uptake value (SUVmax), the contrast ratio (CR) of the SUV to the cerebellum (CR brain), and the CR of the SUV to the contralateral lung (CR lung). The CR lung was calculated from the SUV of the tumor (T) and that of the contralateral normal lung (N) and then was measured by two formulas, namely, T N/T N and T/N. Results. The sensitivities of both CR lung T N/T N and CR lung T/N were significantly higher than those of visual assessment, SUVmax, and CR brain (p 0.01 to p < 0.001). No significant difference in sensitivity was observed between the CR lung T N/T N and CR lung T/N. Both CR lung T N/T N and CR lung T N successfully imaged well-differentiated lung adenocarcinoma more frequently than the visual assessment, SUVmax, and CR brain (p to p < 0.001), whereas there were no significant differences of sensitivity among those five methods for the diagnosis of other histologic types of pulmonary malignancies. Conclusions. The FDG uptake evaluated by the CR lung is superior to that evaluated using the visual assessment, SUVmax, and CR brain for the diagnosis of pulmonary malignancies, especially for well-differentiated lung adenocarcinoma. The simplified formula of CR lung with T/N can be used in place of that with T N/T N. (Ann Thorac Surg 2009;87:886 92) 2009 by The Society of Thoracic Surgeons Recent advances in positron emission tomography (PET) with fluorodeoxyglucose (FDG) have made a useful contribution to the discrimination between benign and malignant nodules [1 6]. The FDG uptake on PET has been evaluated in several ways, such as visual assessment, the standard uptake value (SUV), and the ratio of uptake relative to that of a normal organ. Visual assessment is usually based upon a comparison of FDG uptake by the lesion with normal mediastinal blood pool [7], and is the simplest of the above analyses, although nodules with similar FDG uptake to the mediastinum are difficult to be evaluated. To assess the FDG uptake more objectively, the maximum standardized uptake value (SUVmax) has frequently been used. However, several factors can affect the SUVmax, such as body size [8 10], blood glucose level [11 13], and lesion size [14, 15]. Actually, the mean values of Accepted for publication Sept 30, Address correspondence to Dr Ohba, Department of Thoracic Surgery, Graduate School of Medical Sciences, Kumamoto University, Honjo, Kumamoto, , Japan; oyasumi@kumamoto-u.ac.jp. SUVmax of pulmonary malignancies have been reported to range widely from 5.5 to 10.1 [1 3, 16 18]. There have been reported several ways for evaluating FDG uptake on PET. Nomori and colleagues [19] compared sensitivities and specificities of the SUVmax, the contrast ratio (CR) of the SUV to the lung, and the CR of the SUV to the brain for diagnosing malignant pulmonary nodules, and reported that the CR of the SUV to the lung or brain was more sensitive than the SUVmax for nodules that were faintly positive based on visual findings [19]. The CR of the SUV to the lung in their study was calculated from the SUV of the tumor (T) and that of the contralateral normal lung (N) and then was measured using the formula T N/T N. Their study was further supported in a study by Obrzut and coworkers [20], who reported that the CR of the SUV to the brain was more sensitive than SUVmax for diagnosing malignant pulmonary tumors. The purposes of the present study were as follows: (1) to determine the optimum method of evaluating FDG uptake to discriminate between malignant and benign 2009 by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg OHBA ET AL 2009;87: SEMIQUANTITATIVE ASSESSMENTS OF FDG-PET 887 Abbreviations and Acronyms CR contrast ratio CT computed tomography FDG fluorodeoxyglucose MD moderately differentiated N normal lung PD poorly differentiated PET positron emission tomography ROI region of interest SUV standardized uptake value SUVmax maximum standardized uptake value T tumor WD well differentiated nodules by comparing the visual assessment, SUVmax, the CR to the brain, and the CR to the contralateral lung; (2) to evaluate the utility of the simplified formula of CR to the lung, namely, T/N, by comparing it with that of T N/T N; and (3) to examine the histologic type of malignant nodules with false negative results according to each of the above criteria. In the present study, we targeted malignant tumors less than 3 cm in size because of the following reasons: (1) FDG uptake is dependent on tumor size [15]; and (2) the usefulness of FDG-PET should be examined for pulmonary nodules less than 3 cm because pulmonary masses larger than 3 cm are usually not difficult to diagnose even without FDG-PET. Material and Methods Eligibility The examination of FDG-PET in patients with lung cancer was approved by the Ethical Committee of Kumamoto University Hospital in January Informed consent was obtained from all patients after explaining the costs and benefits of the examinations with their surgeons. Patients Between April 2005 and April 2008, a total of 155 patients with 195 pulmonary nodules less than 3 cm in size that were suspected of being or were diagnosed as pulmonary malignancies, underwent FDG-PET in the department of thoracic surgery of Kumamoto University Hospital before surgery. The size of the nodules was measured on computed tomography (CT) using an electric caliber. Of these, 36 nodules with ground glass opacity images on CT and 29 lesions less than 1 cm in size were excluded, because such lesions are known to be difficult to identify using FDG-PET and therefore are usually out of indication for FDG-PET [5]. The remaining 130 nodules in 107 patients, including 89 nonsmall-cell lung cancers (NSCLC), 12 metastatic lung cancers, and 29 benign nodules, were examined in the present study (Table 1). The histologic type of NSCLC was classified according to the World Health Organization classification [21, 22]. The histologic types of malignant tumors were adenocarcinoma in 69 nodules, squamous cell carcinoma in 17, adenosquamous carcinoma in 3, and metastatic lung cancers in 12. Of the 69 adenocarcinomas, the grade of histologic differentiation was well differentiated (WD) in 51 and moderately differentiated (MD) or poorly differentiated (PD) in 18. Of the 29 benign nodules, 21 were old inflammations, 4 were acute inflammations, and 4 were benign tumors. Of the 21 old inflammatory nodules, 19 were detected simultaneously with NSCLC and were diagnosed clinically as old inflammations without a histologic diagnosis for the following reasons: (1) a review of retrospective chest roentgenograms or CT examinations performed before surgery (mean observation period, months; range, 24 to 97) revealed that the sizes of the lesions had remained unchanged; and (2) postoperative follow-up CT examinations showed the sizes of the lesions had remained unchanged for more than 12 months (mean follow-up period, months; range, 12 to 36). Therefore, the sizes of the 19 old inflammatory nodules had remained unchanged for more than 36 months throughout the preoperative and postoperative periods. The other 2 old inflammatory nodules, 4 acute inflammatory nodules, 4 benign tumors, 89 NSCLC, and 12 metastatic lung cancers were histologically diagnosed from the resected specimens. PET-CT Scanning The PET scanning in all patients was conducted in the Japanese Red Cross Kumamoto Health Care Center in Kumamoto by using an integrated PET/CT device (Discovery ST; GE Medical Systems, Kumamoto, Japan) that consisted of a PET scanner (Advance Nx; GE Medical Systems) and an eight-section CT scanner (Light Speed Plus; GE Medical Systems). Patients were instructed to fast for at least 5 hours before intravenous administration of FDG. Table 1. Characteristics of Pulmonary Lesions Characteristic Value Nonsmall cell lung cancer (n 101) Size (cm) Mean Range 1 3 Histologic subtype WD adenocarcinoma 51 MD or PD adenocarcinoma 18 Squamous cell carcinoma 17 Adenosquamous carcinoma 3 Metastatic lung tumors 12 Benign nodules (n 29) Size (cm) Mean Range 1 3 Histologic type Acute inflammation 4 Old inflammation 21 Benign tumor 4 MD moderately differentiated; PD poorly differentiated; WD well-differentiated.

3 888 OHBA ET AL Ann Thorac Surg SEMIQUANTITATIVE ASSESSMENTS OF FDG-PET 2009;87: The dose of FDG administered was 100 Ci/kg (3.7 MBq/ kg) of body weight. Before the examination of PET, blood sugar level was confirmed to be less than 150 mg/dl in all patients. For patients with the blood sugar level higher than 150 mg/dl, the PET examination was postponed. The PET imaging was performed approximately 60 minutes after intravenous administration of FDG. All images were acquired under shallow-breathing conditions. The acquisition time for PET in three-dimensional mode was 3 minutes per table position. The CT data were resized from a matrix to a matrix to match the PET data to allow image fusion, and a CT transmission map was generated. The PET image data were reconstructed iteratively using the ordered subsets expectation-maximization algorithm with segmented attenuation correction (4 iterations, 28 subsets) and the CT data. The 3.75-mm thick transaxial CT images were reconstructed at 3.27-mm intervals (transaxial) for fusion with the transaxial PET images. The PET, CT, and fused images were available for review in the axial, coronal, and sagittal planes using Xeleris software (GE Medical Systems) on a computer workstation. PET Data Analysis Images were reviewed on a consensus basis by two observers who were unaware of the clinical data. Each observer recorded a visual assessment for each nodule by comparison with FDG uptake on mediastinal blood flow. Lesions with greater FDG uptake than the mediastinal blood flow was defined as positive, and those with less FDG uptake as negative. A consensus was reached if any difference in their opinions existed. The PET data were used to calculate the SUVmax, the CR of the SUV to the contralateral lung (CR lung), and the CR of the SUV to the brain (CR brain). After image reconstruction, a two-dimensional circular region of interest (ROI) was drawn in a slice after visual detection of the highest count on the fused CT images. From these ROI, the maximum activity in the ROI was calculated as lesion activity/injected dose/body weight. The contrast ratio of the SUV between the lesions and the lung (CR lung) and that between the lesions and the brain (CR brain) were then calculated, as described previously [5,19]. Briefly, to calculate the CR brain, the SUVmax in the tumor (T) ROI and the cerebellum (C) were measured, and the CR brain was calculated using the formula T/C, as described previously [19]. To calculate the CR lung, ROIs were placed over the tumor and the contralateral normal lung; then the SUVmax in the tumor (T) ROI and that in the normal lung (N) were measured. The CR lung was then calculated by two kinds of formulas, namely, T N/T N and T/N. Finally, the visual assessment, SUVmax, CR lung T N/T N, CR lung T/N, and CR brain were compared with each other. Determining Cutoff Value of Each Criterion A receiver operating characteristic curve was constructed according to each criterion using SPSS software (SPSS Fig 1. The receiver operating characteristics curve for discriminating nonsmall cell lung cancer/benign nodules. (A) Maximum standardized uptake value (SUVmax); (B) contrast ratio (CR) brain; (C) CR lung T N/T N; (D) CR lung T/N. (T tumor; N normal lung.)

4 Ann Thorac Surg OHBA ET AL 2009;87: SEMIQUANTITATIVE ASSESSMENTS OF FDG-PET 889 Table 2. Summary of Results of SUVmax, CR Brain, CR Lung T N/T N, and CR Lung T/N Diagnosis Criterion NSCLC Benign Total Sensitivity Specificity Visual estimation Positive Negative SUVmax CR brain CR lung T N/ T N a CR lung T/N b Total a The sensitivity of the formula CR lung T N/T N is significantly higher than that of visual estimation, SUVmax, and CR brain (McNemar test: p 0.01, p 0.001, and p 0.001). b The sensitivity of the formula CR lung T/N is significantly higher than that of visual estimation, SUVmax, and CR brain (McNemar test: p 0.002, p 0.001, and p 0.001). CR contrast ratio; N normal lung; SUVmax maximum standardized uptake value; T tumor J for Windows; SPSS, Chicago, IL), and the cutoff values were determined for benign/malignant discrimination. Nodules with more than the cutoff value of FDG uptake were defined as positive on FDG-PET. Statistical Analysis The Fisher exact test was used to compare the distribution of size between malignant and benign nodules. True positive, true negative, false positive, and false negative results of each criterion for detecting NSCLC were compared with the pathologic diagnosis. Sensitivity was calculated as [true positive/true positive false negative], and specificity as [true negative/true negative false positive], and the differences among the criteria were analyzed using the McNemar test. Statistical analysis was performed using SPSS software. All values in the text and tables are given as mean SD. Results The receiver operating characteristic curve showed that the optimal cutoff values for SUVmax, CR brain, CR lung T N/T N, and CR lung T/N were 1.1, 0.16, 0.29, and 1.83, respectively (Fig 1A through D). Therefore, nodules with an SUVmax of 1.1 or greater, CR brain of 0.16 or greater, CR lung T N/T N of 0.29 or more, and CR lung T/N of 1.83 or more were defined as positive in each criterion, whereas nodules with lower values were defined as negative. Table 2 summarizes the sensitivity and specificity for discriminating malignant/benign nodules by each criterion. The sensitivities for diagnosis of NSCLC were 0.78 for visual assessment, 0.74 for SUVmax, 0.73 for CR brain, 0.89 for CR lung T N/T N, and 0.91 for CR lung T/N. Although the sensitivities of both the CR lung T N/T N and CR lung T/N were significantly higher than those of visual assessment (p 0.01 and p 0.002, respectively), SUVmax (p for both), and CR brain (p for both), there was no significant difference in sensitivity among the visual assessment, the SUVmax, and the CR brain, or between the CR lung T N/T N and the CR lung T/N. There was no significant difference of specificity among these five criteria. Table 3 shows the number of tumors with true positive and false negative according to each criterion for WD adenocarcinomas (n 51), MD or PD adenocarcinomas (n 18), and nonadenocarcinomas including metastatic lung cancers (n 32). Of the 51 WD adenocarcinomas, false negative results were shown in 20 (39%) for visual assessment, 22 (43%) for SUVmax, 21 (41%) for CR brain, 10 (20%) for CR lung T N/T N, and 8 (16%) for CR lung T/N. None of the MD or PD adenocarcinomas showed false negative results for each criterion, except for 1 MD adenocarcinoma evaluated using CR brain. All of the nodules with false negative results in nonadenocarcinomas were metastatic lung cancers. Table 4 summarizes the sensitivity according to each Table 3. Results of Each Criterion in Well-Differentiated (WD) Adenocarcinoma, Moderately Differentiated (MD) or Poorly Differentiated (PD) Adenocarcinoma, and Nonadenocarcinoma WD Adenocarcinoma MD or PD Adenocarcinoma Nonadenocarcinoma Criterion Positive Negative Positive Negative Positive Negative Visual estimation SUVmax CR brain CR lung T N/T N CR lung T/N CR contrast ratio; N normal lung; SUVmax maximum standardized uptake value; T tumor.

5 890 OHBA ET AL Ann Thorac Surg SEMIQUANTITATIVE ASSESSMENTS OF FDG-PET 2009;87: Table 4. Summary of Sensitivity of Each Criterion in Well-Differentiated (WD) Adenocarcinoma, Moderately Differentiated (MD) or Poorly Differentiated (PD) Adenocarcinoma, and Nonadenocarcinoma Sensitivity Criterion WD Adenocarcinoma MD or PD Adenocarcinoma Nondenocarcinoma Visual estimation SUVmax CR brain CR lung T N/T N 0.80 a CR lung T/N 0.84 b a The sensitivity for WD adenocarcinoma of the formula CR lung T N/T N is significantly higher than that of visual estimation (p 0.002), SUVmax (p 0.001), and CR brain (p 0.002). b The sensitivity for WD adenocarcinoma of the formula CR lung T/N is significantly higher than that of visual estimation (p 0.001), SUVmax (p 0.001), and CR brain (p 0.001). CR contrast ratio; N normal lung; SUVmax maximum standardized uptake value; T tumor. criterion for WD adenocarcinoma, MD or PD adenocarcinoma, and nonadenocarcinoma. For WD adenocarcinoma, CR lung T N/T N had a significantly higher sensitivity than visual assessment, SUVmax, and CR brain (p 0.002, p 0.001, and p 0.002, respectively); CR lung T/N also had a significantly higher sensitivity than visual assessment, SUVmax, and CR brain (p 0.001). There was no significant difference in sensitivity for the diagnosis of MD or PD adenocarcinoma or nonadenocarcinoma among those five criteria. Comment The present study showed the following two points: (1) FDG uptake measured by CR lung has higher sensitivity for diagnosis of pulmonary malignancy as compared with the visual assessment, SUVmax, and CR brain, especially for WD adenocarcinoma; and (2) the simplified formula of CR lung calculated by T/N showed similar sensitivity to CR lung T N/T N, even for WD adenocarcinoma. Although SUVmax has been frequently used for the semiquantitative analysis of FDG uptake, it has been reported that several factors can affect the SUV, such as body size [8 10], blood glucose level [11 13], and lesion size [14, 15]. In breast cancers, Wahl and colleagues [23] have demonstrated that a ratio of SUV between the tumor and contralateral normal breast tissue was more reliable than the absolute SUV of tumor for diagnosing breast malignancies. Nomori and coworkers [5, 24] have used a CR lung calculated using the formula T N/T Nto diagnose pulmonary malignancies; this formula is according to that used in the study on breast cancer [25]. In the present study, to simplify the formula of T N/T N, we evaluated CR lung using the formula of T/N. As a result, both the CR lung T N/T N and CR lung T/N showed higher sensitivity than visual assessment, SUVmax, and CR brain, and there was no significant difference between the CR lung T N/T N and CR lung T/N. Therefore, we conclude that both formulas of CR lung are superior to SUVmax and CR brain for diagnosis, and also that the simple formula of CR lung T/N can be used in place of CR lung T N/T N. The present study showed that most of the malignant tumors with false negative results by each criterion were WD lung adenocarcinomas. It has been reported that FDG- PET gives false negative results for low-grade lung cancer, such as bronchioloalveolar carcinoma and WD adenocarcinoma, because of the low glucose metabolism and low tumor cell density [5, 6, 26]. In the present study, both the CR lung T N/T N and the CR lung T/N showed higher sensitivity for the diagnosis of WD adenocarcinoma than the visual assessment, SUVmax, and CR brain, whereas there was no significant difference in sensitivity for the diagnosis of MD or PD adenocarcinoma or nonadenocarcinoma among those five criteria. Therefore, the FDG uptake of WD adenocarcinoma should be evaluated using the CR lung rather than the visual assessment, SUVmax, or CR brain. While SUVmax with a cutoff value of 2.5 has been frequently used as a criterion for diagnosing pulmonary malignancies using FDG-PET [27], the cutoff values of SUVmax in the present study was much lower, namely, 1.1. This difference could be due to the following: (1) whereas a cutoff value of 2.5 for SUVmax has been used to diagnose pulmonary tumors, including tumors larger than 3 cm, the present study restricted the size of pulmonary nodules to between 1 cm and 3 cm; and (2) the 69 of 89 NSCLC (78%) examined in the present study were adenocarcinomas, which are known to usually show a lower FDG uptake than nonadenocarcinoma NSCLC [5, 19, 26]. We conclude that FDG uptake evaluated by CR lung is superior to that evaluated by the visual assessment, SUVmax, or CR brain for the diagnosis of pulmonary malignancies, especially for WD adenocarcinoma. Because there was no significant difference between CR lung T N/T N and CR lung T/N, the simpler latter formula can be used in place of the former for the diagnosis of pulmonary malignancies. References 1. Scott WJ, Schwabe JL, Gupta NC, et al. Positron emission tomography of lung tumors and mediastinal lymph nodes using [18F] fluorodeoxyglucose. Ann Thorac Surg 1994;58:

6 Ann Thorac Surg OHBA ET AL 2009;87: SEMIQUANTITATIVE ASSESSMENTS OF FDG-PET Patz EF, Lowe VJ, Hoffman JM, et al. Focal pulmonary abnormalities: evaluation with F-18 fluorodeoxyglucose PET scanning. Radiology 1993;188: Gupta NC, Maloof J, Gunel E. Probability of malignancy in solitary pulmonary nodules using fluorine-18-fdg and PET. J Nucl Med 1996;37: Gambhir SS, Shepherd JE, Shah BD, et al. Analytical decision model for the cost-effectiveness management of solitary pulmonary nodules. J Clin Oncol 1998;16: Nomori H, Watanabe K, Ohtsuka T, Naruke T, Suemasu K, Uno K. Evaluation of F-18 fluorodeoxyglucose (FDG) PET scanning for pulmonary nodules less than 3 cm in diameter, with special reference to the CT images. Lung Cancer 2004;45: Nomori H, Watanabe K, Ohtsuka T, Naruke T, Suemasu K, Uno K. The size of metastatic foci and lymph nodes yielding false-negative and false-positive lymph node staging with positron emission tomography in patients with lung cancer. J Thorac Cardiovasc Surg 2004;127: Lowe VJ, Hoffman JM, DeLong DM, Patz EF, Coleman RE. Semiquantitative and visual analysis of FDG-PET images in pulmonary abnormalities. J Nucl Med 1994;35: Zasadny KR, Wahl RL. Standarized uptake values of normal tissues at PET with 2-[fluorine-18]-fluoro-2-deoxy-Dglucose: variations with body weight and a method for correction. Radiology 1993;189: Kim CK, Gupta NC, Chandramouli B, Alavi A. Standarized uptake values of FDG: body surface area correction is preferable to body weight correction. J Nucl Med 1994;35: Kim CK, Gupta NC. Dependency of standardized uptake values of fluorine-18 fluorodeoxyglucose on body size: comparison of body surface area correction and lean body mass correction. Nucl Med Commun 1996;17: Lindholm P, Minn H, Leskinen-Kallio S, Bergman J, Ruotsalainen U, Joensuu H. Influence of the blood glucose concentration on FDG uptake in cancer: a PET study. J Nucl Med 1993;34: Langen KJ, Braun U, Rota Kops E, et al. The influence of plasma glucose levels of fluorine-18-fluorodeoxyglucose uptake in bronchial carcinomas. J Nucl Med 1993;34: Hamberg LM, Hunter GI, Alpert NM, Choi NC, Babich JW, Fischman AJ. The dose uptake ratio as an index of glucose metabolixm: useful parameter or oversimplification? J Nucl Med 1994;35: Cremerius U, Fabry U, Neuerburg J, Zimny M, Osieka R, Buell U. Positron emission tomography with 18F-FDG to detect residual disease after therapy for malignant lymphoma. Nucl Med Commun 1998;19: Menda Y, Bushnell DL, Madsen MT, McLaughlin K, Kahn D, Kernstine KH. Evaluation of various corrections to the standardized uptake value for diagnosis of pulmonary malignancy. Nucl Med Commun 2001;22: Dewan NA, Gupta NC, Redepenning LS, et al. Diagnostic efficacy of PET-FDG imaging in solitary pulmonary nodules. Potential role in evaluation and management. Chest 1993; 104: Imdahl A, Jenkner S, Brink I, et al. Validation of FDG positron emission tomography for differentiation of unknown pulmonary lesions. Eur J Cardiothorac Surg 2001;20: Lowe VJ, Fletcher JW, Gobar L, et al. Prospective investigation of positron emission tomography in lung nodules. J Clin Oncol 1998;16: Nomori H, Watanabe K, Ohtsuka T, Naruke T, Suemasu K, Uno K. Visual and semiquantitative analyses for F-18 fluorodeoxyglucose (FDG) PET scanning in pulmonary nodules 1 to 3 cm in size. Ann Thorac Surg 2005,79: Obrzut S, Pham RH, Vera DR, Badran K, Hoha CK. Comparison of lesion-to-cerebellum uptake ratios and standardized uptake values in the evaluation of lung nodules with 18F-FDG PET. Nucl Med 2007;27: Travis M, Colvy T, Corrin B, et al. World Health Organization international histological classification of tumors. Histological typing of lung and pleural tumors. 3rd ed. Berlin: Springer-Verlag, Sobin LW, Witteking CH. UICC TMN classification of malignant tumors. 6th ed. New York: Wiley-Liss 2002: Wahl RL, Cody RL, Hutchins GD, et al. Primary and metastatic breast carcinoma: initial clinical evaluation with PET with the radiolabeled glucose analogue 2-[F-18]-fluoro-2- deoxy-d-glucose. Radiology 1991;179: Watanabe K, Nomori H, Ohtsuka T, et al. [F-18] Fluorodeoxyglucose positron emission tomography can predict pathological tumor stage and proliferative activity determined by Ki-67 in clinical stage IA lung adenocarcinomas. Jpn J Clin Oncol 2006;36: Uno K, Yoshikawa K, Imazeki K, Okada J, Minoshima S, Arimizu N. FDG-PET evaluation of breast carcer: a diagnostic role for primary or metastatic and recurrent lesions. Radiology 1992;185: Higashi K, Ueda Y, Seki H, et al. Fluorine-18-FDG PET imaging is negative in bronchioloalveolar carcinoma. J Nucl Med 1998;39: Coleman RE. PET in lung cancer. J Nucl Med 1999;40: INVITED COMMENTARY In the 1990s, fluorodeoxyglucose positron emission tomography (FDG-PET) was introduced as a part of the evaluation of suspicious solitary pulmonary nodules, and visual assessment and the maximum standard uptake value (SUVmax) pixel in the region of interest have been used to estimate the likelihood of malignancy [1, 2]. Most studies have concentrated on larger lesions of 3 cm or more [3]. FDG-PET is unlikely to be as accurate for the smaller lesions. So, the question asked by Ohba and colleagues [4] can the accuracy of FDG-PET be improved in evaluating malignancy in lesions less than 3 cm? is certainly relevant. Others have attempted to improve the accuracy of PET by manipulating the emission results [2, 5]. Ohba and colleagues, like Nomori and colleagues [6], derived their calculations without demonstrating the physiologic reason why the calculations might be better than visual assessment or SUVmax to assess the solitary pulmonary nodules. The PET data manipulations may not be generalizable to other patient groups or FDG-PET systems: the machines, FDG injected, the FDG-PET protocols, analytic software, and the readers differ across institutions. There are no consistencies. There appears to be several areas of concern in which bias may have been introduced: 1. FDG-PET may not be applicable to patients who are diabetic, especially those who have poor glucose control. This article has no exclusions for high glucose levels. 2. The authors excluded 23% (36 of 154) of the solitary pulmonary nodules because of ground glass opacity by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur