Extrapulmonary benign and malignant lesions avid for 18F-fluoro-deoxyglucose by multivariate regression model identification

Extrapulmonary benign and malignant lesions avid for 18F-fluoro-deoxyglucose by multivariate regression model identification 1 2 corresp first-author Corresponding author. E-mail address: fudanzhsh@yahoo.com.cn Corresponding author. E-mail address: fudanzhsh@yahoo.com.cn fudanzhsh@yahoo.com.cn 1 2 Department of Nuclear Medicine in the First Affiliated Hospital of Jinan University, Guangzhou 510632, China Department of Nuclear Medicine in the First Affiliated Hospital of Jinan University, Guangzhou 510632, China PET/CT Center in the First Affiliated Hospital of Guangzhou Medical College, Guangzhou 510230, China PET/CT Center in the First Affiliated Hospital of Guangzhou Medical College, Guangzhou 510230, China Whether extrapulmonary lesions are avid for 18 F-fluorodeoxyglucose ( 18 F-FDG) could help to differentiate the benign or malignant lung lesions. In this trial, the 199 consecutive patients with newly diagnosed lung lesions (169 malignant and 36 benign lesions) were imaged by whole body 18 F-FDG PET/CT. Histopathology and clinic results served as the reference standard. The malignancy likelihood were conducted by CT scores ; the maximum standardized uptake value ( SUV max ) of lung lesions, and PET on FDG negative or positive, as well as metastasis index ( MI ), by PET combined with CT findings. The data were analyzed by stepwise logistic regression and receiver-operating- characteristic. The malignancy predictive probability ( P ) was obtained by P =e x /(1+e x ), where x = –1.16+0.87 ( CT score ) +0.15( SUV max )+0.27( MI ). The area under curve (AUC) for the fitted logistic model was 0.82±0.04, this was superior and significantly different from SUV max (AUC, 0.73±0.05) and CT scores (AUC, 0.71±0.05). The fitted logistic model could improve the diagnostic performance. The MI could help for differential diagnosis. Lung neoplasm 18 F-fluorodeoxyglucose Positron-emission tomography and computed tomography Logistic models Receiver-operating-characteristic 1 Introduction 1 As the differential diagnosis of pulmonary nodules and mass lesions in many countries, the positron emission tomography (PET) with 18 F-fluorodeoxyglucose ( 18 F- FDG) has been established. According to a meta- analysis of the published data on 18 F-FDG–PET scanning from January 1996 to September 2000, the average sensitivity and specificity of 18 F-FDG–PET scanning for detecting a malignancy was 97% and 78% [ 1 1 ] . The 18 F-FDG–PET has not the benefit of evaluating pulmonary nodules with low 18 F-FDG avid [ 2 2 ] . Compared with 18 F-FDG–PET, the combination of 18 F-FDG–PET with the diagnostic CT scan without intravenous contrast can improve the sensitivity from 69% to 97%, and does not change the specificity [ 3 3 ] . The diagnostic accuracy of 18 F-FDG PET/CT without the quality CT of lung nodule was similar with that of 18 F-FDG–PET [ 4 4 , 5 5 ] . Nie et al. [ 6 6 ] reported a semiautomatic computer-assisted diagnostic (CAD) scheme, indicating that the combination of 18 F- FDG–PET with CT can differentiate benign from malignant pulmonary nodules, and is better than 18 F- FDG–PET or CT alone. In that study, we excluded two nodules with extrathoracic malignancy, and focused on lung lesions in the whole body 18 F-FDG–PET. Lesion outside of lung cancer avid for 18 F-FDG is considered as metastasis. The extrapulmonary lesion is avid for 18 F-FDG, indicating that the newly diagnosed lung lesion is malignant. 2 Methods and materials 1 2.1 Patient cohort 2 With a lung lesion newly diagnosed at the conventional chest radiography or the X-ray CT for a suspicious malignant primary, the patients underwent whole body integrated 18 F-FDG PET/CT from February 2005 to January 2009. We selected the pathological confirmation and suspicious infection patients cured by antibiotic, and excluded the patients with recent malignancy history and diabetes mellitus. We gained the approval of Institutional Review Boards without patient informed consent. The 199 consecutive patients (63 women and 136 men, 25–88 years old, mean age± SD : 58.8±12.4) met the further analysis criteria. In 205 lung lesion patients, their biopsy or surgery proved that the 169 was malignant; and 32, benign in pathology; and 4 infection cured by antibiotics (Table 1). 1 1 Malignants 1 1 Numbers 1 1 Benigns 1 1 Numbers 1 1 Adenocarcinoma 1 1 101 1 1 Tuberculosis 1 1 11 1 1 Squamous cell carcinoma 1 1 34 1 1 Sarcoidosis 1 1 2 1 1 Adenosquamous carcinoma 1 1 4 1 1 Organized pneumonia 1 1 4 1 1 Large cell carcinoma 1 1 1 1 1 Abscess 1 1 4 1 1 Small cell carcinoma 1 1 10 1 1 Cryptococcal pneumonitis 1 1 1 1 1 Synoviosarcoma 1 1 1 1 1 Infection/inflammation 1 1 10 1 1 Hodgkin's lymphoma 1 1 1 1 2 Other benign lesions were not further classified 1 2 4 1 1 Metastatic carcinoma from rectum/ lung 1 1 3 1 1 Unspecified NSCLC 1 1 14 1 1 / 1 1 / 1 1 Total number 1 1 169 1 1 / 1 1 36 2.2 Acquisition of PET/CT image 2 To fulfill the blood glucose level of less than the 8.3 mmol/dL, patients were asked to fast over 6 h and rest for 15 min before administration 5-MBq/kg 18 F-FDG. The images were acquired by a whole-body PET/CT scanner (Discovery DST 8; GE Healthcare, USA) after injection at 60 min. The whole body scan was approximately from the middle thighs to the skull roof. The CT scan without intravenous contrast was used as a protocol involving 140 kV, 150 mA, 0.8 s per tube-rotation, and 3.75-mm slice thickness. The PET scan was performed with a 3.27-mm section thickness, 3.5 min per table position, and two-dimensional acquisitions. Patients were asked to maintain normal shallow respiration during image acquisition. The 18 F-FDG-PET images were reconstructed using CT attenuation correction and an ordered subset expectation maximization algorithm. The maximum standardized uptake value ( SUV max ) was determined by Ref.[ 7 7 ]. 2.3 Acquisition of CT images 2 All patients underwent a breath-hold spiral CT scan on the lung lesions without intravenous contrast by 120kV, 170 mA, 0.8 second per tube-rotation, 2.5-mm slice thickness, and 1.35-pitch after PET/CT scan. 2.4 Interpretation of CT images 2 The CT images were interpreted by two radiologists, who were unaware of each patient's history and PET images. Each lesion was described by its site, size, attenuation, shape, margin characteristics, consolidation, cavitation, and invasion. The criteria for interpreting lesions were applied [ 5 5 ] . Readings in case of differing results were performed in consensus. 2.5 Interpretation of PET/CT images 2 The 18 F-FDG–PET criteria for malignancy were used as citation [ 8 8 ] . The locations of abnormal tracer uptake were recorded, and the metastasis index ( MI ) was scored by a 5-points scale ( Table 2 Table 2 ). The PET/CT images were analyzed by two doctors with experience of 5 years in the 18 F-FDG PET/CT. Also readings were conducted in consensus in case of differing results. 2.6 Statistical analysis 2 The patient characteristics were compared by the Student’s t -test and the chi-squared test. When the expected values in the any cells of the contingency table were below five, the Fisher exact test was conducted. Variables reaching the significance might be included in a multivariate logistic regression model. Selection method with entry testing was based on the significance of the score statistic; and removal testing, the probability ( P > 0.15) of the likelihood ratio test [ 9 9 ] . Odds ratios ( OR ) and the 95% CIs were computed by unconditional logistic regression. A receiver-operating -characteristic (ROC) analysis included the CT interpretations, and the lesion uptakes in SUV max , and the predicted malignancy probability ( P ) which was calculated by the fitted multivariate logistic regression model. The difference in the area under curve (AUC) was tested by the Z statistic. A 2-tailed p values (≤ 0.05) showed statistical significant differences. The diagnostic OR for a test is defined as sensitivity/(1‒ sensitivity) ×specificity/(1‒specificity) [ 10 10 ] . 3 Results 1 3.1 Results of CT and PET/CT 2 The patient characteristics and their lesions are shown in Table 2 Table 2 . In the 58 lesions, their lesion sizes could hardly be determined because of the local atelectasis and its associated obstruction, and parahilar location. In the 147 lesions, the lesion sizes were 36±21 mm in the range of 4–143 mm). Locations of distant lesions avid for 18 F-FDG are listed in Table 3 Table 3 . The non-contrast CT shows that 62% (104/169) of malignant lesions were classified as positive; and 29% (59/205) of all lesions, equivocal; and 44% (16/36) of benign lesions, negative ( Table 2 Table 2 ). Taking equivocal lesion as malignancy, the diagnostic OR was 8.2; and as benignancy, 3.6. In Table 2 Table 2 , the sensitivity for 18 F-FDG PET/CT was 95% (161/169) and the specificity was 25% (9/36) at the diagnostic OR of 6.7. The sex, age, and lesion sizes were no obvious difference between benign and malignant lesions, but the CT score , SUV max , 18 F- FDG score , and MI were highly different ( p < 0.01). a It was hard to measure the 58 lesions. b The lesions avid for 18 F-FDG in the lung hilar nodes was scored as 1; and ipsilateral mediastinal nodes, as 2; and contralateral mediastinal nodes, as 3; and distant locations, as 4. There was not any suspicious metastasis for the scored as 0. 1 1 Characteristic 1 1 1 1 Malignancy 1 1 Benignancy 1 1 p 1 1 Patients 1 1 Women/Men 1 1 50/113 1 1 13/23 1 1 0.58 1 1   1 1 Age / year 1 1 59±13 1 1 57±12 1 1 0.29 1 1 Lesions 1 1 Size (mm) a 1 1 37±22( n =122) 1 1 31±20 ( n =25) 1 1 0.21 1 1   1 1 CTscore (0/1/2) 1 1 15/50/104 1 1 16/9/11 1 1 <0.01 1 1   1 1 18F-FDGscore(0/1) 1 1 8/161 1 1 9/27 1 1 <0.01 1 1   1 1 SUVmax 1 1 10.7±6.0 1 1 5.7±3.7 1 1 <0.01 1 1   1 1 MI(0/1/2/3/4) b 1 1 48/11/34/15/61 1 1 21/1/7/3/4 1 1 <0.01 1 1 Location/number 1 1 Malignancy 1 1 Benignancy 1 1 Celiac lymph nodes 1 1 8 1 1 2 1 1 Retroperitoneal lymph node 1 1 8 1 1 1 1 1 Pelvic lymph nodes 1 1 2 1 1 1 1 1 Inguinal lymph nodes 1 1 1 1 1 1 1 1 Lymph node of neck 1 1 12 1 1 / 1 1 Axillary nodes 1 1 4 1 1 / 1 1 Tonsil 1 1 1 1 1 / 1 1 Skeleton 1 1 26 1 1 1 1 1 Liver 1 1 5 1 1 1 1 1 Suprarenal gland 1 1 7 1 1 / 1 1 Brain 1 1 3 1 1 / 1 1 Cervical cord 1 1 1 1 1 / 1 1 Thyroid gland 1 1 / 1 1 1 1 1 Abdominal wall 1 1 1 1 1 / 1 1 Subcutaneous soft tissue 1 1 2 1 1 / 1 1 Gluteus 1 1 1 1 1 / 1 1 Pulmonary trunk and left atrium 1 1 1 1 1 / 1 1 Total number 1 1 83 1 1 7 3.2 Multivariate stepwise logistic regression model analysis 2 By stepwise selection, CT score , SUV max , and MI were included in the fitted multivariate logistic regression model without 18 F -FDG score . In this model, when increasing, either CT scores or SUV max or MI was used as malignancy predictors. The P was calculated by P =e x / (1+e x ), where x = –1.16+0.87( CT score )+0.15 ( SUV max ) + 0.27( MI ) ( Table 4 Table 4 ). 1 1 Variables 1 1 Coefficient ( β ± SD ) 1 1 Odds Ratio (95% confidence interval) 1 1 p 1 1 CT score (0/1/2) 1 1 0.87±0.27 1 1 2.39(1.41–4.05) 1 1 <0.01 1 1 SUV max 1 1 0.15±0.05 1 1 1.16(1.04–1.29) 1 1 <0.01 1 1 MI (0/1/2/3/4) 1 1 0.27±0.14 1 1 1.31(1.00–1.73) 1 1 0.05 1 1 Constant 1 1 –1.16±0.47 1 1 0.31(0.12–0.79) 1 1 0.01 The model estimated that the P was more than 50% at x ≥0. The CT score calculated by the equation was 2 at x > 0. In other words, at the positive CT scan, the lung lesion should be taken as malignancy regardless of 18 F-FDG uptake [ 3 3 ] . For example, an adenocarcinoma in the right upper lobe was noted, its CT score was 2; and SUV max , 1.5; and MI , 0; and P , 69.1%, as shown in Fig.1 Fig.1 . At the same time, when CT score was 1 and SUV max was not less than 2, the x was larger than zero. To put it at the indeterminate CT scan, the SUV max =2.0 was used as cutoff for malignancy [ 3 3 , 11 11 ] , thus helping to distinguish between benign and malignant lesions. When CT score was zero, the x was larger than zero at the SUV max of larger than 7.7, suggesting that the lung lesion should be considered as malignancy regardless of its morphological information. When diagnostic results of CT score were contradictive with the SUV max , the MI could help for differential diagnosis. For example, an inflammation lesion in the left lower lobe was observed, its CT score was zero; and SUV max , 5.0; and MI , zero; and P , 39.9% (Fig.2). 3.3 ROC analysis 2 A ROC analysis for CT score, SUV max and P was performed ( Fig.3 Fig.3 ). The SUV max curve based on the lung lesions has the cutoff of initial 1.5, step 0.5, and end 8.0. The SUV max curve intersects the CT score curve. The additional values on CT score and SUV max curves hold over the entire range of sensitivity and specificity. The AUC was 0.83±0.04 for the model; and 0.71±0.05, for CT score ; and 0.74±0.05, for SUV max. Statistical analysis by the AUC shows that the model was superior and significantly different from SUV max ( p =0.01) and CT score ( p < 0.01). There was no significant difference between SUV max and CT score at p =0.64. At the cutoff value of 0.50 for malignancy, the sensitivity was 96% (163/169); and specificity, 31% (11/36); and the diagnostic OR for P , 12.0. 4 Discussion 1 In the 18 F-FDG–PET, the SUV max at OR =1.16 ( p < 0.01) was associated with lung malignancy in the model ( Table 4 Table 4 ). The SUV max increased as a significant malignancy predictor was similar to Grgic’s report [ 10 10 ] . The sensitivity of 18 F-FDG PET/CT scanning for detecting malignancy was 95%, this was equivalent to the available data of 97% sensitivity by meta- analysis [ 1 1 ] , and a prospective multicenter study of 92% sensitivity [ 8 8 ] . But the 25% specificity in our study was lower than those of the reported 44%–85% [ 1 1 , 3 3 , 4 4 , 5 5 , 8 8 , 11 11 , 12 12 , 13 13 ] . Also, the benign lesions, including tuberculosis, bacterial pneumonia, organized pneumonia, active sarcoidosis, infectious granulomas, and acute pyogenic abscesses, and so forth, can produce false-positive readings in 18 F-FDG–PET [ 14 14 ] , these were mainly caused by the low specificity. The CT morphological information was an effective predictor for malignancy at OR =2.41( p < 0.01), and its sensitivity and specificity were not as accurate in this study as in a multicenter of contrast-enhanced CT (98% sensitivity, and 58% specificity) [ 15 15 ] . As we know, a lung lesion at CT diagnosis suspicious for a malignant primary is usually dependent on morphological information and enhancement [ 16 16 ] . Some equivocal lesions may shift to definitely malignant or definitely benign on its enhancement, and improve the CT sensitivity and specificity. The less accuracy was likely caused without contrast-enhanced CT [ 15 15 ] . The MI was included in the model at OR =1.31( p =0.05). To the best of our knowledge, the MI is the first proof whether lesions outside of lungs avid for 18 F-FDG might help to distinguish the malignant lung lesion. Based on the score statistic probability ( p =0.05), we should be cautious to make a conclusion by MI . The 18 F- FDG score was excluded from the model because its information had been included in its SUV max . The stepwise logistic regression was designed to find the most parsimonious set of predictors in predicting malignancy. The ROCs show that the model lies over the curve for CT diagnosis and SUV max , indicating that the model accuracy is superior regardless of where setting the threshold defining a malignancy test. Furthermore, the statistical test by the AUC shows that the superior model was different from CT score and SUV max . Compared with the diagnostic OR of 3.7 or 8.3 in different diagnosis protocol for CT and 18 F-FDG (6.3), the diagnostic OR reached the highest value of 12.0 at the cutoff of 0.50 P , suggesting the model was better than CT and SUV max alone. The model had potential to develop a semiautomatic CAD to aid and improve physician diagnostic skills, and it would be tested by a multicenter in the future. 5 Limitation 1 In this retrospective clinical study, our patients were selected from their larger pool, and referred to the 18 F-FDG PET/CT study because of the pathologic verification of their lung lesions, these induced a selection bias. The malignancy rate was 82% (169/205). No matter how there were several influencing factors such as serum glucose level, respiration, partial-volume effects, and noises [ 17 17 ] , the SUV max was in use as the de facto standard [18] . 6 Conclusion 1 In this study, the extrapulmonary lesions avid 18 F-FDG could distinguish the malignancy from benignancy. Our logistic model including CT score , SUV max , and MI information for predicting the malignancy P was superior to CT and 18 F-FDG alone, thus improving physician diagnostic performance. References Gould M K , Maclean C C , Kuschner WG , et al . JAMA , 2001 , 285 : 914 - 924 . JAMA Kim S K , Allen-Auerbach M , Goldin J , et al . J Nucl Med , 2007 , 48 : 214 - 220 . J Nucl Med Chen Y C H , Chen, P , Tian J H , et al . Nucl Sci Tech , 2009 , 20 : 17 - 21 . Nucl Sci Tech Pauls S , Buck A K , Halter G , et al . Mol Imaging Biol , 2008 , 10 : 121 - 128 . Mol Imaging Biol Nie Y , Li Q , Li F , et al . J Nucl Med , 2006 , 47 : 1075 - 1080 . J Nucl Med Boellaard R , O'Doherty M.J , Weber W.A , et al . Eur J Nucl Med Mol Imaging , 2010 , 37 : 181 - 200 . Eur J Nucl Med Mol Imaging Fletcher J W , Kymes S M , Gould M , et al . J Nucl Med , 2008 , 49 : 179 - 185 . J Nucl Med Hosmer D W , Lemeshow S . Applied logistic regression . 2nd edition . New Jersey : John Wiley & Sons, Inc , 2000 , 31 - 142 . Applied logistic regression Eadie, L.H , Taylor P , Gibson A . Eur J Radiol , 2012 , 81 : e70 - e76 . Eur J Radiol Grgic A , Yüksell Y , Gröschel A , et al . Eur J Nucl Med Mol Imaging , 2010 , 37 : 1087 - 1094 . Eur J Nucl Med Mol Imaging Suga K , Kawakami Y , Hiyama A , et al . Ann Nucl Med , 2009 , 23 : 427 - 435 . Ann Nucl Med Tian J H , Yang X F , Yu LJ , et al . J Nucl Med , 2008 , 49 : 186 - 194 . J Nucl Med The American College of Radiology . Acr practice guideline for performing FDG-PET/CT in oncology. ACR. 2007 . http://www.acr.org/SecondaryMainMenuCategories/quality_safety/guidelines/nuc_med/fdg_pet_ct.aspx http://www.acr.org/SecondaryMainMenuCategories/quality_safety/guidelines/nuc_med/fdg_pet_ct.aspx ( June 18, 2009 ). Acr practice guideline for performing FDG-PET/CT in oncology. ACR. 2007 Swensen S J , Viggiano RW , Midthun D E , et al . Radiology , 2000 , 214 : 73 - 80 . Radiology Lee K.S , Yi C A , Jeong S Y , et al . Chest , 2007 , 131 : 1516 - 1525 . Chest Basu S , Zaidi H , Holm S , et al . Curr Med Imaging Rev , 2011 , 7 : 216 - 233 . Curr Med Imaging Rev Wahl R L , Jacene H , Kasamon Y , et al . J Nucl Med , 2009 , 50 : 122S - 150S . J Nucl Med We thank Dr Xin Cai in the PET center of the First Affiliated Hospital of Guangzhou Medical College in China who had contributed to get CT images interpretation to us. 10.13538/j.1001-8042/nst.2013.04.007 Supported by National Natural Science Foundation of China [Grant No. 30800274] 2012-10-20 2013-08-01 2021-04-11 2013 Nuclear Science and Techniques 4 24