| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 All authors: Department of Radiology, Mayo Clinic, 200 First St. SW, Rochester, MN 55905.
Received July 6, 2005;
accepted after revision November 11, 2005.
Address correspondence to J. A. Christensen
(christensen.jared{at}mayo.edu).
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Examinations from patients undergoing both nodule-enhancement CT and 18F-FDG PET to characterize the same indeterminate SPN were reviewed. For nodule-enhancement CT, an SPN was considered malignant when it showed an unenhanced to peak contrast-enhanced increase in attenuation greater than 15 H. Fluourine-18-FDG PET studies were blindly reinterpreted by two qualified nuclear radiologists. SPNs qualitatively showing hypermetabolic activity greater than the mediastinal blood pool were interpreted as malignant. These interpretations were compared with the original prospective clinical readings and to semiquantitative standardized uptake value (SUV) analysis. Results were compared with pathologic and clinical follow-up.
RESULTS. Forty-two pulmonary nodules were examined. Twenty-five (60%) were malignant, and 17 (40%) were benign. Nodule-enhancement CT was positive in all 25 malignant nodules and in 12 benign nodules, with sensitivity and specificity of 100% and 29%, respectively, and with a positive predictive value (PPV) and negative predictive value (NPV) of 68% and 100%, respectively. Qualitative 18F-FDG PET interpretations were positive in 24 of the 25 malignant nodules and in four benign nodules. Fluourine-18-FDG PET was considered negative in one malignant nodule and in 13 of the 17 benign nodules. This correlates with a sensitivity and specificity of 96% and 76%, respectively, and with a PPV and NPV of 86% and 93%, respectively. Original prospective 18F-FDG PET and semiquantitative SUV analysis showed sensitivity, specificity, PPV, and NPV of 88%, 76%, 85%, and 81% and 84%, 82%, 88%, and 78%, respectively.
CONCLUSION. Due to its much higher specificity and only slightly reduced sensitivity, 18F-FDG PET is preferable to nodule-enhancement CT in evaluating indeterminate pulmonary nodules. However, nodule-enhancement CT remains useful due to its high NPV, convenience, and lower cost. Qualitative 18F-FDG PET interpretation provided the best balance of sensitivity and specificity when compared with original prospective interpretation or SUV analysis.
Keywords: chest • CT imaging • lung diseases • PET/CT
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Two techniques currently used to characterize indeterminate SPNs are nodule-enhancement CT and 18F-FDG PET. Nodule-enhancement CT is performed under the premise that a neoplastic lesion, with its increased vascularity, will enhance when imaged with IV contrast material. Lesions that enhance greater than 15 H from the unenhanced level to peak contrast-enhancement are considered likely malignant, whereas those that enhance less than 15 H are considered likely benign. A recent multicenter analysis of nodule-enhancement CT using these criteria showed a sensitivity of 98% and a specificity of 58% [10]. Fluourine-18-FDG PET uses 18F-FDG as a marker of metabolism with lesions localizing 18F-FDG proportionate to their metabolic activity. SPNs with hypermetabolism greater than the mediastinal blood pool are likely malignant. To further quantify metabolism in individual nodules, a standardized uptake value (SUV) can be calculated [11]. An SUV greater than 2.5 defines the SPN as malignant with a relatively high degree of sensitivity and specificity [11]. Using these criteria, recent studies have shown a sensitivity of 92-96% and a specificity of 77-90% using 18F-FDG PET [11, 12].
|
|
|
|
|
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nodule-enhancement CT was performed on each patient using HiSpeed Advantage, HiSpeed CT/i, or LightSpeed Qx/i helical scanners (GE Healthcare). A standardized protocol was used in the imaging of each patient using a collimation of 3 mm with a narrow field of view including the nodule. Images were obtained before and after administration of IV contrast material (420 mg I/kg injected at 300 mg/mL) at 1, 2, 3, and 4 minutes. The original prospective nodule-enhancement CT examination reports were retrospectively reviewed. Examination reports that did not state the attenuation difference were reviewed and the attenuation difference was measured by an experienced radiologist. An increase in attenuation within the nodule from unenhanced to peak contrast enhancement greater than 15 H was considered to be malignant. An increase from unenhanced to peak contrast enhancement less than 15 H was interpreted as benign (Figs. 1A, 1B, 2A, and 2B).
Fluourine-18-FDG PET was performed using an Advance PET or a Discovery LS PET/CT scanner (GE Healthcare). Whole-body attenuation-corrected images were obtained from the base of the skull through the symphysis pubis after administering a standard dose of 15 mCi of 18F-FDG IV and allowing for a 60-minute equilibration. Each patient had been fasting at least 4 hours when imaged.
Fluourine-18-FDG PET examinations were interpreted three ways to test for differences in sensitivity and specificity using both qualitative and semiquantitative interpretations. First, the original 18F-FDG PET reports were reviewed to determine the original prospective characterization of each SPN as benign or malignant. Second, a blinded qualitative reinterpretation of each 18F-FDG PET examination was performed independently by two experienced nuclear radiologists. In this qualitative reinterpretation, a nodule was defined as malignant if it had hypermetabolic activity greater than the metabolic activity of the adjacent mediastinal blood pool (Figs. 3A and 3B). Nodules with no activity or activity less than that of the adjacent mediastinal blood pool were identified as benign. Last, a semiquantitative measure of metabolism was performed for each SPN. A region of interest was placed over each nodule and average SUVs were determined using body-weight normalization. Nodules with an SUV greater than or equal to 2.5 were considered malignant. Nodules with an SUV less than 2.5 were considered benign.
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nodule-enhancement CT was interpreted as positive (enhancement > 15 H) in all 25 malignant nodules and in 12 of the benign nodules. Five nodules were interpreted as negative (enhancement < 15 H) on nodule-enhancement CT. All five of these nodules were benign. These numbers represent a cumulative sensitivity of 100% and a specificity of 29%, with a positive predictive value (PPV) and negative predictive value (NPV) of 68% and 100%, respectively. The 12 false-positive SPNs for malignancy on nodule-enhancement CT were as follows: five nodules exhibited radiologic stability or regression (two by follow-up chest X-ray and three by CT). The other seven nodules showed benign histology after open resection or biopsy: one nodule each as histoplasmosis, chondroid hamartoma, sarcoidosis, foreign body reaction to talc, and caseating granuloma; and two nodules were interpreted as benign histology not otherwise specified (Figs. 4A, 4B, and 4C).
Prospective 18F-FDG PET was interpreted as positive (original report interpreted as positive for malignancy by the primary interpreting radiologist) in 22 of the 25 malignant nodules and in four benign nodules. Sixteen nodules were considered benign on prospective 18F-FDG PET. Thirteen of these were benign. This represents a sensitivity of 88% and a specificity of 76%, with a PPV and NPV of 85% and 81%, respectively. The four false-positive SPNs were one each of histoplasmosis, sarcoidosis, foreign body reaction to talc, and caseating granuloma. Two small grade 2 adenocarcinomas (13 x 8 x 7 mm in the right lower lobe and 7 x 7 x 5 mm in the left upper lobe) and one atypical carcinoid were the three malignant nodules initially interpreted as negative for malignancy by the primary radiologist (Figs. 5A and 5B).
Blinded, qualitative retrospective 18F-FDG PET interpretation (using visual analysis: metabolic activity in the nodule greater than the metabolic activity of the adjacent mediastinal blood pool) was positive in 24 of the 25 malignant nodules and in four benign nodules. Fourteen nodules were interpreted as benign on 18F-FDG PET. Thirteen of these were defined as benign by imaging follow-up or pathology. This represents a sensitivity of 96% and a specificity of 76%, with a PPV and NPV of 86% and 93%, respectively. The four false-positive nodules were one each of histoplasmosis, sarcoidosis, foreign body reaction to talc, and benign histology not otherwise specified. The single false-negative nodule was one of the same grade 2 adenocarcinomas interpreted as false-negative by the prospective clinical 18F-FDG PET interpretation (7 x 7 x 5 mm in the left upper lobe). There was no interobserver variability between the two blinded, independent 18F-FDG PET reviewers.
Semiquantitative 18F-FDG PET was interpreted as positive (SUV
2.5) in 21 of the 25 malignant nodules and in three of the benign
nodules. Eighteen nodules were considered benign by SUV criteria (SUV <
2.5). Fourteen of these were truly benign. This represents a sensitivity of
84% and a specificity of 82%, with a PPV and NPV of 88% and 78%, respectively.
The three false-positives included one each of histoplasmosis, sarcoidosis,
and foreign body reaction to talc. Three of the four nodules interpreted as
false-negative for malignancy were the same atypical carcinoid and the same
two grade 2 adenocarcinomas interpreted as negative by the prospective
radiologist. The remaining false-negative nodule was a 10 x 15 mm grade
2 adenocarcinoma interpreted as malignant by both prospective and blinded
qualitative 18F-FDG PET.
|
|
Four nodules were imaged initially with 18F-FDG PET, followed by nodule-enhancement CT. All of these nodules were benign. Three had positive results on the initial prospective 18F-FDG PET examination. One of these three nodules was negative on subsequent nodule-enhancement CT. This nodule was also negative and interpreted as benign on retrospective qualitative and semiquantitative 18F-FDG PET interpretation. The benign nodule that was interpreted as negative on initial prospective 18F-FDG PET was also interpreted as negative on retrospective qualitative and semiquantitative 18F-FDG PET but was interpreted as positive on subsequent nodule-enhancement CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Nodule-enhancement CT is one such technique. A multicenter study performed by Swensen et al. [10] showed a sensitivity of 98% and a 96% NPV. The same study, however, also showed a relatively low specificity of only 58% and a PPV of 68%. The high sensitivity and high NPV of nodule-enhancement CT coupled with its relatively low cost and high general availability render it a useful examination in SPN characterization. However, the poor specificity of this examination can lead to increased overall costs and greater morbidity due to unnecessary biopsies and other thoracic surgical interventions.
Fluorine-18-FDG PET has also been widely studied for its use in the accurate characterization of SPNs. A multicenter study by Lowe et al. [11] showed sensitivities of 100% and 80% and specificities of 74% and 95% for visual and SUV analyses of SPNs, respectively. A recent meta-analysis of studies using 18F-FDG PET for SPN evaluation showed a sensitivity of 96.8% and a specificity of 77.8% [12]. Although a more costly examination than nodule-enhancement CT, the higher specificity of 18F-FDG PET can ultimately lead to considerable cost savings by reducing the number of biopsies and surgical interventions. Prior limited access to 18F-FDG PET is rapidly resolving because the number of both fixed and mobile 18F-FDG PET units has increased significantly during the past 5 years. However, one of the major limitations is the resolution of the scanners used. Currently, the resolving limit of 18F-FDG PET scanners is nodules that are 6-8 mm in size. Nodules smaller than this are not adequately evaluated with 18F-FDG PET.
Although studies have evaluated the individual potential of nodule-enhancement CT and 18F-FDG PET to accurately characterize an SPN as benign or malignant, to date there has not been a study directly comparing the two techniques. The results from our comparison of 18F-FDG PET and nodule-enhancement CT indicate that, in support of prior studies, nodule-enhancement CT shows a very high sensitivity and NPV (both 100%) but a very poor specificity (29%). Fluorine-18-FDG PET gives a slightly lower sensitivity yet much higher specificity and PPV than nodule-enhancement CT. In addition, our results show that the best balance of sensitivity and specificity for malignancy is obtained when 18F-FDG PET images are interpreted by qualitatively comparing the metabolic activity within the SPN to the metabolic activity of the adjacent mediastinal blood pool. A degree of SPN activity greater than that of the mediastinal blood pool is highly suggestive of malignancy, with a sensitivity of 96% and a specificity of 76%. Our study showed a slightly higher specificity of 83% when semiquantitatively qualifying an SPN using an SUV threshold of 2.5. However, we found a lower sensitivity (88%) for malignancy when solely using SUV measurements. Overall, our study supports the use of 18F-FDG PET over nodule-enhancement CT in the evaluation of indeterminate SPNs due to its high sensitivity and much better specificity.
The major limitation of our study is referral bias. The majority of SPNs (38 of 42 [90%]) were imaged with nodule-enhancement CT first. With a low pretest probability for malignancy, a positive nodule-enhancement CT for a given SPN may have been considered equivocal by the referring physician, dictating subsequent referral for 18F-FDG PET and a higher incidence of malignancy in our study population. This would also certainly have the potential to increase the false-positive rate for nodule-enhancement CT and reduce its overall specificity. Also, due to the very high NPV of nodule-enhancement CT, many nodules not demonstrating enhancement on nodule-enhancement CT may have been considered benign and thus not referred for 18F-FDG PET. This reduction in potential true-negative nodule-enhancement CT cases from the study population could lead to a reduction in nodule-enhancement CT specificity and again, a higher incidence of malignancy in the study population. Despite these referral bias limitations, the difference in specificity between 18F-FDG PET and nodule-enhancement CT in our study is so great that our data continue to support the preferential use of 18F-FDG PET for evaluation of SPNs, particularly when there is a high index of suspicion for malignancy. In fact, the 76% specificity for qualitative 18F-FDG PET interpretation is in line with multiple prior studies [8, 11, 12, 14, 15] and well above the 58% specificity established for nodule-enhancement CT in the large multicenter trial published by Swensen et al. [10].
Several nodules in our study were characterized erroneously by both techniques. In the nodule-enhancement CT group, there were no false-negative nodules. There were three false-negative nodules in the 18F-FDG PET group. These occurred in two low-grade adenocarcinomas and in an atypical carcinoid. Encountering these false-negative nodules was not surprising, considering the reliance of 18F-FDG PET on metabolic activity. It has previously been shown that 18F-FDG PET can misdiagnose malignant lesions with low metabolic activity, such as is seen in low-grade neoplasms and, classically, in slow-growing carcinoid tumors [16, 17]. Our results are consistent with this. However, the majority of tumors are of higher metabolic activity, and 18F-FDG PET, although misdiagnosing some low-grade malignant nodules, still performed quite well in our study, correctly diagnosing the majority of malignant nodules.
There were several false-positive results in the nodule-enhancement CT group, which lowered the overall specificity considerably. The false-positive results occurred in several nodules that were benign histologically or that proved to be benign after stability or regression on radiologic follow-up. These histologically confirmed false-positive nodules included histoplasmosis, chondroid hamartoma, sarcoidosis, foreign body reaction to talc, and necrotizing granuloma, all conditions with inflammatory components and thus increased vascularity. Nodules observed to be false-positive in the 18F-FDG PET examinations included histoplasmosis, sarcoidosis, foreign body reaction to talc, caseating granuloma, and nonspecific benign abnormality at resection. The inflammation inherent to the majority of these false-positive nodules would certainly lead to increased metabolic activity. It is understandable that the impact of these false-positive and false-negative nodule-enhancement CT and 18F-FDG PET SPN examinations on patient management could be lessened by correlating the imaging findings with other important clinical parameters such as history, physical examination, risk factors, and laboratory studies.
Due to the high sensitivity and NPV of nodule-enhancement CT, nodules with a negative nodule-enhancement CT have an extremely low likelihood of being malignant. Patients with a negative nodule-enhancement CT thus would not benefit from further 18F-FDG PET examination. However, nodule-enhancement CT generates a large number of false-positives owing to its low specificity. In cases where an SPN is positive on nodule-enhancement CT, subsequent 18F-FDG PET examination would likely be helpful. In our study, nine of the 38 nodules (24%) examined initially with nodule-enhancement CT were erroneously interpreted as possibly malignant. Eight of these were subsequently interpreted as benign on 18F-FDG PET. Relying on the nodule-enhancement CT results alone would have subjected these eight patients with benign nodules to costly interventions in which potential morbidity is relatively high, such as open thoracotomy, video-assisted thoracoscopic surgery, or transthoracic needle aspiration. Indeed, these results show the added value of performing 18F-FDG PET after a positive nodule-enhancement CT study in an indeterminate SPN.
In conclusion, our recommendations for the use of nodule-enhancement CT and 18F-FDG PET in the evaluation of an indeterminate SPN greater than 7 mm are summarized in algorithmic form in Figure 6. We believe that 18F-FDG PET, with its high sensitivity and much higher specificity than nodule-enhancement CT, is the superior technique for accurately characterizing indeterminate SPNs. However, the lower sensitivity of 18F-FDG PET does have the capacity of generating a greater number of false-negative results, especially in lower grade malignancies. The lower sensitivity of 18F-FDG PET combined with the higher cost must be factored into the evaluation of any pulmonary nodule.
With its high sensitivity and NPV, widespread availability, and relatively low cost, nodule-enhancement CT could continue to be used as an integral part of the evaluation of indeterminate SPNs in patients with a low likelihood of malignancy. At the initial discovery of an SPN, all patients should undergo a thin-section CT examination to accurately characterize the nodule. Nodules in patients with a low likelihood of malignancy could undergo nodule-enhancement CT, preferably during the same examination. Nodules with less than 15 H of enhancement on nodule-enhancement CT have a very low probability of malignancy and can be defined as benign, with no further imaging or follow-up needed. Nodules exhibiting greater than 15 H of enhancement on nodule-enhancement CT, however, should undergo subsequent 18F-FDG PET examination. These patients benefit from the higher specificity of 18F-FDG PET, eliminating the false-positive results generated from the low specificity of nodule-enhancement CT.
Patients with a moderate or high likelihood of malignancy should undergo 18F-FDG PET evaluation. In our review, it appears that a nonbiased, qualitative interpretation of 18F-FDG PET through comparison of SPN and mediastinal blood pool metabolic activity provides the best balance of sensitivity and specificity for the accurate diagnosis of malignant nodules. SUV analysis can then be reserved for those patients in whom the qualitative 18F-FDG PET interpretation is equivocal. In our study, three malignant nodules interpreted as malignant by qualitative visual interpretation were considered negative by SUV analysis because the SUV fell below 2.5 (SUV of 1.7, 1.8, and 1.7) (Fig. 7). Nodules with metabolic activity greater than the mediastinal blood pool are likely malignant and should undergo further invasive resection or biopsy. Those with metabolic activity less than the mediastinal blood pool are likely benign. However, due to the imperfect sensitivity of 18F-FDG PET, we would recommend that these nodules be examined with serial radiologic imaging for further workup of malignant conditions.
A recent study by Comber et al. [7] in an Australian setting supports these recommendations. This study showed a diagnostic strategy using both nodule-enhancement CT and 18F-FDG PET to be more cost-effective than 18F-FDG PET with conventional CT alone (incremental cost:accuracy ratio of Australian $12,059.18:Australian $12,636.36). This remained the case except in populations with a high disease prevalence, where the cost-effectiveness of both strategies was shown to be approximately equal [7]. By coupling the high sensitivity of nodule-enhancement CT with the much higher specificity and overall greater accuracy of 18F-FDG PET, clinicians can accurately characterize indeterminate SPNs as benign or malignant, thereby reducing the number of invasive procedures required for diagnosis as well as providing for the early determination of malignancy while definitive surgical resection is still feasible and potentially curative.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  C.-J. Chen, B.-F. Lee, W.-J. Yao, L. Cheng, P.-S. Wu, C. L. Chu, and N.-T. Chiu Dual-Phase 18F-FDG PET in the Diagnosis of Pulmonary Nodules with an Initial Standard Uptake Value Less Than 2.5 Am. J. Roentgenol., August 1, 2008; 191(2): 475 - 479. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Zou, M. Zhang, Q. Wang, D. Shang, L. Wang, and G. Yu Quantitative Investigation of Solitary Pulmonary Nodules: Dynamic Contrast-Enhanced MRI and Histopathologic Analysis Am. J. Roentgenol., July 1, 2008; 191(1): 252 - 259. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Y. Donmez, E. Yekeler, V. Saeidi, A. Tunaci, M. Tunaci, and G. Acunas Dynamic Contrast Enhancement Patterns of Solitary Pulmonary Nodules on 3D Gradient-Recalled Echo MRI Am. J. Roentgenol., December 1, 2007; 189(6): 1380 - 1386. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. A. Margolis, J. M. Hoffman, R. J. Herfkens, R. B. Jeffrey, A. Quon, and S. S. Gambhir Molecular Imaging Techniques in Body Imaging Radiology, November 1, 2007; 245(2): 333 - 356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Veronesi, M. Bellomi, U. Veronesi, G. Paganelli, P. Maisonneuve, P. Scanagatta, F. Leo, G. Pelosi, L. Travaini, C. Rampinelli, et al. Role of Positron Emission Tomography Scanning in the Management of Lung Nodules Detected at Baseline Computed Tomography Screening Ann. Thorac. Surg., September 1, 2007; 84(3): 959 - 966. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
