Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print January 30, 2008, 10.1148/radiol.2463062148
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow All Versions of this Article:
2463062148v1
246/3/772    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cronin, P.
Right arrow Articles by Carlos, R. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cronin, P.
Right arrow Articles by Carlos, R. C.
(Radiology 2008;246:772-782.)
© RSNA, 2008

Evidence-based Practice

Solitary Pulmonary Nodules: Meta-analytic Comparison of Cross-sectional Imaging Modalities for Diagnosis of Malignancy1

Paul Cronin, MD, MS, Ben A. Dwamena, MD, Aine Marie Kelly, MD, MS, and Ruth C. Carlos, MD, MS

1 From the Department of Radiology, Divisions of Cardiothoracic Radiology (P.C., A.M.K.), Nuclear Medicine (B.A.D.), and Magnetic Resonance Imaging (R.C.C.), University of Michigan Medical Center, B1 132F Taubman Center/0302, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0030; and Veterans Administration Ann Arbor Healthcare System Nuclear Medicine Service, Ann Arbor, Mich (B.A.D.). Received December 20, 2006; revision requested February 13, 2007; revision received May 23; accepted June 13; final version accepted August 29. Funded in part by National Institutes of Health/National Cancer Institute grant 1 K07 CA108664 01A1 and a GE-AUR Radiology Research Academic Fellowship. Address correspondence to P.C. (e-mail: pcronin{at}med.umich.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Purpose: To perform a meta-analysis to estimate the diagnostic accuracy of dynamic contrast material–enhanced computed tomography (CT) and magnetic resonance (MR) imaging, fluorine 18 fluorodeoxyglucose (FDG) positron emission tomography (PET), and technetium 99m (99mTc) depreotide single photon emission computed tomography (SPECT) for evaluation of solitary pulmonary nodules (SPNs).

Materials and Methods: Data sources were studies published in PubMed between January 1990 and December 2005. The selected investigations were comparative and noncomparative diagnostic cohort studies to examine the operating characteristics of the four imaging modalities for evaluation of SPNs, involving at least 10 enrolled participants with histologic confirmation and having sufficient data to calculate contingency tables. A random coefficient binary regression model with disease probability conditioned on test results was used to summarize test performance and construct summary receiver operating characteristic (ROC) curves. Sensitivities, specificities, predictive values, diagnostic odds ratios, and areas under the ROC curve were calculated.

Results: Forty-four studies—10 dynamic CT, six dynamic MR, 22 FDG PET, and seven 99mTc-depreotide SPECT—met the inclusion criteria. (One study was included in both the FDG PET and SPECT groups.) Sensitivities, specificities, positive predictive values, negative predictive values, diagnostic odds ratios, and areas under the ROC curve were, respectively, 0.93 (95% confidence interval [CI]: 0.88, 0.97), 0.76 (95% CI: 0.68, 0.97), 0.80 (95% CI: 0.74, 0.86), 0.95 (95% CI: 0.93, 0.98), 39.91 (95% CI: 1.21, 81.04), and 0.93 (95% CI: 0.81, 0.97) for dynamic CT; 0.94 (95% CI: 0.91, 0.97), 0.79 (95% CI: 0.73, 0.86), 0.86 (95% CI: 0.83, 0.89), 0.93 (95% CI: 0.90, 0.96), 60.59 (95% CI: 5.56, 115.62), and 0.94 (95% CI: 0.83, 0.98) for dynamic MR; 0.95 (95% CI: 0.93, 0.98), 0.82 (95% CI: 0.77, 0.88), 0.91 (95% CI: 0.88, 0.93), 0.90 (95% CI: 0.85, 0.94), 97.31 (95% CI: 6.26, 188.37), and 0.94 (95% CI: 0.83, 0.98) for FDG PET; and 0.95 (95% CI: 0.93, 0.97), 0.82 (95% CI: 0.78, 0.85), 0.90 (95% CI: 0.83, 0.97), 0.91 (95% CI: 0.84, 0.98), 84.50 (95% CI: 34.28, 134.73), and 0.94 (95% CI: 0.83, 0.98) for 99mTc-depreotide SPECT.

Conclusion: Dynamic CT and MR, FDG PET, and 99mTc-depreotide SPECT are noninvasive and accurate in distinguishing malignant from benign SPNs; differences among these tests are nonsignificant.

Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/2463062148/DC1

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
The most common manifestation of lung cancer is a solitary pulmonary nodule smaller than 3 cm in diameter, which is usually found during computed tomography (CT), or a solitary pulmonary mass larger than 3 cm in diameter. Diagnostic evaluation of focal pulmonary lesions should be accurate and efficient to facilitate prompt resection of malignant tumors, when possible, but surgery should be avoided in cases of benign disease (1,2).

It has been hypothesized that benign and malignant pulmonary lesions demonstrate distinctly different physiologic responses (3), including a greater degree of contrast enhancement of malignant solitary pulmonary nodules. Since the early 1990s, this hypothesis has been applied to CT, magnetic resonance (MR) imaging, positron emission tomography (PET), and more recently single photon emission computed tomography (SPECT) (4).

Fluorine 18 fluorodeoxyglucose (FDG) PET is approximately 90% accurate in the characterization of malignant versus benign solitary pulmonary nodules, with 97% sensitivity and 78% specificity (5). More recently, the results of examinations performed with technetium 99m (99mTc) depreotide, a somatostatin analogue, suggest that these tests have sensitivity and specificity similar to those of FDG PET (6). This radionuclide is imaged with a 99mTc-depreotide SPECT scanner, and 99mTc-depreotide SPECT is a promising alternative for centers that do not offer PET. Unlike PET, this noninvasive functional imaging test is based on the principle that malignant nodules have a higher level of somatostatin-avid receptors than do benign nodules.

The tissue contrast resolution achieved with MR imaging is generally superior to that achieved with CT, and, thus, MR imaging might serve as a complementary modality for differentiating solitary pulmonary nodules by depicting the internal composition of the nodules. To date, the role of MR imaging in the clinical management of patients who have lung cancer or focal and diffuse lung diseases has been limited because of certain MR characteristics of the lungs, such as low proton density, numerous air-tissue interfaces, and respiratory or cardiac motion artifacts (7). The purpose of this study was to perform a meta-analysis to estimate the diagnostic accuracy of dynamic contrast material–enhanced CT and MR imaging, FDG PET, and 99mTc-depreotide SPECT for the evaluation of solitary pulmonary nodules.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
We used published methods to identify relevant studies, assess study eligibility, evaluate the methodologic quality of the studies (8–10), and summarize the diagnostic accuracy findings (8–11).

Data Sources
We initiated this quantitative systematic review by performing a comprehensive computer search of the English-language medical literature or English-language abstracted studies by using primarily the prespecified filter for diagnostic performance in the PubMed (MEDLINE) database to identify original peer-reviewed, full-length articles on the evaluation of solitary pulmonary nodules in humans (9). The following key words were used: computed tomography, CT, dynamic contrast enhanced computed tomography, DCE CT, magnetic resonance imaging, MRI, dynamic contrast enhanced magnetic resonance imaging, DCE MRI, positron emission tomography, PET, fluoro-2-deoxy-D-glucose positron emission tomography, FDG PET, single photon emission computed tomography, 99mTc-depreotide single photon emission computed tomography, 99mTc-depreotide SPECT, solitary pulmonary nodule, SPN, solitary lung nodule, coin lesion, and diagnostic test. No attempt was made to include unpublished data.

Study Identification
We identified all studies published between January 1990 and December 2005 in which dynamic contrast-enhanced CT (or dynamic CT), dynamic contrast-enhanced MR imaging (or dynamic MR imaging), FDG PET, or functional 99mTc-depreotide SPECT was evaluated for assessment of solitary pulmonary nodules. Considering the time that the clinical use of each modality was introduced, the searches extended from January 1992 to December 2005 for dynamic CT, from January 1995 to December 2005 for dynamic MR imaging, from January 1990 to December 2005 for FDG PET, and from January 1999 to December 2005 for 99mTc-depreotide SPECT.

Study Eligibility
One investigator (B.A.D.) evaluated the English-language studies for potential inclusion and reviewed the articles to determine their eligibility for detailed analysis. The articles selected for inclusion and analysis met the following criteria: The diagnostic performance of dynamic CT, dynamic MR imaging, FDG PET, or 99mTc-depreotide SPECT in the evaluation of solitary pulmonary nodules and in the differentiation between malignant and benign pulmonary nodules and mass lesions was assessed; imaging results were compared with histologic sample (percutaneous or surgical biopsy, or surgical resection) findings for more than 50% of the patients; results were reported in sufficient detail to reconstruct contingency tables of the raw data (ie, true-positive, true-negative, false-positive, and false-negative findings); there was a minimal sample size of 10 patients; and established diagnostic criteria were used to identify abnormal test results.

Data Abstraction
Each of two investigators (B.A.D., P.C.) abstracted the following information from all of the eligible articles without blinding: author(s), study location, journal in which the study was published, year of publication, study design, number of patients, demographic characteristics of participants, mean size of solitary pulmonary nodules, and reference-standard examination used. Each investigator independently abstracted data regarding the number of patients with and the number of patients without disease and the number of true-positive, true-negative, false-positive, and false-negative imaging results.

Study Quality
Each of three reviewers (P.C., R.C.C., B.A.D.) independently assessed the quality of each study according to prospectively developed criteria that were modified from well-accepted methodologic standards for evaluating quality in diagnostic test research; disagreements were resolved by means of discussion and consensus (8,10). The following nine criteria were evaluated, and a grade of 1 was given for each criterion that was met: prospective study design, sample size of 30 or more subjects, adequate description and quality of the index test, appropriate reference-standard test used, complete verification of results with use of the reference-standard test, adequate clinical description of subjects, adequate reporting of results, broad population, and blinded interpretation of test results.

Data Synthesis and Statistical Analyses
Diagnostic test characteristics.—For each study, sensitivities, specificities, positive predictive values (PPVs), negative predictive values (NPVs), and diagnostic odds ratios—with relevant 95% confidence intervals (CIs)—were recalculated from the true-positive, true-negative, false-positive, and false-negative results in the contingency tables.

Meta-analysis model.—Test performance was summarized by using a random coefficient binary regression model (12–14). This model, formulated such that probability of disease was conditioned on test results, represents a clinically sensible approach to the meta-analysis of diagnostic test accuracy studies, allowing fixed and random-effects modeling of clinically relevant, individualizable measures of diagnostic test performance and estimation of summary receiver operating characteristic (ROC) values. The expected operating points on the summary ROC curves were constructed by choosing a clinically plausible range of false-positive rates (1 – specificity) and using the estimated parameters from the binary regression model to compute the predicted true-positive rates (sensitivity). As a function of the mixed-effects model, the data were intrinsically weighted by using the individual study variances.

We introduced random-effects terms for the intercepts and slopes to account for between- and within-study variability. The logistic regression model was extended by including covariate factors as explanatory variables in the linear part of the model. All estimations were performed by using the GLLAMM (generalized linear latent and mixed models) module in Stata software (Stata, College Station, Tex).

Quality assessment and study heterogeneity.—Separate analyses, with covariates entered as fixed effects, were performed for potentially relevant study design and methodologic quality indexes, which were entered as dichotomous variables (ie, yes = 1, no = 0) such as clinically relevant population, adequate description of study subjects, satisfactory description of index test, consecutive patient selection, sample size of 30 or more subjects, prospective data collection, use of the appropriate reference-standard examination, complete verification of test results, blinded interpretation of test results, and satisfactory reporting of results. Summary statistics are presented. Since heterogeneity was expected a priori, the degree of inconsistency across studies was measured by using the I2 test, which describes the proportion of variation in treatment effect estimates that is due to genuine variation rather than to sampling error (15).

Testing for publication bias.—Publication bias is the systematic error introduced by publication status. For example, studies that reveal an effect may be more likely to be published than studies that show no effect. The normal quantile plot, which is used to compare the quantiles of an observed distribution against the quantiles of the standard normal distribution (16), was used to test the following null hypotheses: (a) The effect-size estimates reported in the included studies meet the normality assumption, (b) the effect-size estimates are derived from a single population, and (c) there is no evidence of publication bias in the included literature. Forest plots were used to visualize the extent of heterogeneity among studies.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Study Identification
The search initially yielded 242 potential literature citations (Fig 1). One hundred eighty-one of these studies were excluded because they were not relevant, because they were not clinical trials, or because of the electronic limiting of the search to humans, English-language medical literature, or English-language abstracted studies. At review of the abstracts from the remaining 61 articles, an additional 12 trials were excluded; reasons included the use of a different modality, such as chest radiography, for the comparisons; the use of different contrast agents such as non–commercially available radioisotopes; and the discovery that one study was a meta-analysis. After this exclusion, 49 studies were left for full publication review. At review of the complete articles, five more studies were excluded because of the lack of histologic information or the lack of sufficient information to complete a contingency table. After this final screening, 44 published trials met our inclusion criteria (6,9–55). One study had an English-language abstract, with the full article published in Japanese (45). In one study (53), the efficiency of both FDG PET and 99mTc-depreotide SPECT in the diagnosis of malignant solitary pulmonary nodules was assessed.


Figure 1
View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Flowchart illustrates the selection of studies. In this article, the efficiency of both FDG PET and 99mTc-depreotide SPECT in the diagnosis of malignant solitary pulmonary nodules was assessed (*). DCE = dynamic contrast enhanced.

 
Study Characteristics
The 44 evaluated studies involved 2867 patients with 2896 nodules. The trials were published between 1990 and 2005. Twenty-four trials were prospective (Table E1, http://radiology.rsnajnls.org/cgi/content/full/2463062148/DC1). Twenty trials were performed in the United States; 10 trials, in Europe; and 14 trials, in Asia. The average patient age across all studies was 59.5 years. In two studies, the mean age was not specified. The average size of the nodules was 18.6 mm. In four studies, the mean nodule size was not specified. The average number of nodules imaged per study was 65.8. In 19 studies, the final diagnosis for all subjects was proved histologically. In the remaining studies, the final diagnosis was proved by using histologic analysis for some subjects and clinical follow-up for others.

Diagnostic Test Performance Indexes and Summary Estimates
For the 10 dynamic CT studies, the pooled sensitivity was 0.93 (95% CI: 0.88, 0.97); the pooled specificity, 0.76 (95% CI: 0.68, 0.97); the pooled PPV, 0.80 (95% CI: 0.74, 0.86); the pooled NPV, 0.95 (95% CI: 0.93, 0.98); the pooled diagnostic odds ratio, 39.91 (95% CI: 1.21, 81.04) (Table E2, http://radiology.rsnajnls.org/cgi/content/full/2463062148/DC1) (Fig 2); and the area under the ROC curve, 0.93 (95% CI: 0.81, 0.97).


Figure 2
View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Sensitivity, specificity, PPV, and NPV data ({blacksquare}), with corresponding 95% CIs, for dynamic CT detection of malignant solitary pulmonary nodules. The size of the square is related to the degree of variance in the given study. The vertical broken line represents pooled sensitivity, specificity, PPV, and NPV values, and the distortion of the diamond represents 95% CIs of the pooled results.

 
For the six dynamic MR studies, the pooled sensitivity was 0.94 (95% CI: 0.91, 0.97); the pooled specificity, 0.79 (95% CI: 0.73, 0.86); the pooled PPV, 0.86 (95% CI: 0.83, 0.89); the pooled NPV, 0.93 (95% CI: 0.90, 0.96); the pooled diagnostic odds ratio, 60.59 (95% CI: 5.56, 115.62) (Table E2, http://radiology.rsnajnls.org/cgi/content/full/2463062148/DC1) (Fig 3); and the area under the ROC curve, 0.94 (95% CI: 0.83, 0.98).


Figure 3
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Sensitivity, specificity, PPV, and NPV data ({blacksquare}), with corresponding 95% CIs, for dynamic MR detection of malignant solitary pulmonary nodules. The size of the square is related to the degree of variance in the given study. The vertical broken line represents pooled sensitivity, specificity, PPV, and NPV values, and the distortion of the diamond represents 95% CIs of the pooled results.

 
For the 22 FDG PET studies, the pooled sensitivity was 0.95 (95% CI: 0.93, 0.98); the pooled specificity, 0.82 (95% CI: 0.77, 0.88); the pooled PPV, 0.91 (95% CI: 0.88, 0.93); the pooled NPV, 0.90 (95% CI: 0.85, 0.94); the pooled diagnostic odds ratio, 97.31 (95% CI: 6.26, 188.37) (Table E2, http://radiology.rsnajnls.org/cgi/content/full/2463062148/DC1) (Fig 4); and the area under the ROC curve, 0.94 (95% CI: 0.83, 0.98).


Figure 4
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Sensitivity, specificity, PPV, and NPV data ({blacksquare}), with corresponding 95% CIs, for FDG PET detection of malignant solitary pulmonary nodules. The size of the square is related to the degree of variance in the given study. The vertical broken line represents pooled sensitivity, specificity, PPV, and NPV values, and the distortion of the diamond represents 95% CIs of the pooled results.

 
For the seven 99mTc-depreotide SPECT studies, the pooled sensitivity was 0.95 (95% CI: 0.93, 0.97); the pooled specificity, 0.82 (95% CI: 0.78, 0.85); the pooled PPV, 0.90 (95% CI: 0.83, 0.97); the pooled NPV, 0.91 (95% CI: 0.84, 0.98); the pooled diagnostic odds ratio, 84.50 (95% CI: 34.28, 134.73) (Table E2, http://radiology.rsnajnls.org/cgi/content/full/2463062148/DC1) (Fig 5); and the area under the ROC curve, 0.94 (95% CI: 0.83, 0.98). Summary ROC curves (Fig 6) showed no significant differences in values among the four tests.


Figure 5
View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: Sensitivity, specificity, PPV, and NPV data ({blacksquare}), with corresponding 95% CIs, for 99mTc-depreotide SPECT detection of malignant solitary pulmonary nodules. The size of the square is related to the degree of variance in the given study. The vertical broken line represents pooled sensitivity, specificity, PPV, and NPV values, and the distortion of the diamond represents 95% CIs of the pooled results.

 

Figure 6
View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6: Areas under summary ROC curves for dynamic contrast-enhanced CT (DCECT), dynamic contrast-enhanced MR imaging (DCEMRI), FDG PET, and 99mTc-depreotide SPECT.

 
Study Quality Scores and Study Heterogeneity
With regard to the 10 dynamic CT trials, study quality scores ranged from 5 to 9, with 9 being the highest score. Three (30%) CT trials were performed prospectively; eight (80%) trials involved 30 or more subjects; in three (30%) trials, sufficient uniformity of the verification test was reported; in all 10 (100%) trials, a sufficient description of the index test was provided; and nine (90%) trials involved the appropriate reference-standard test. Sufficient information on the patients' characteristics was reported in all 10 (100%) trials, and sufficient information on the diagnostic test characteristics was reported in two (20%). Patients who broadly reflected the general population were recruited in all 10 (100%) trials, and blinded interpretation of test results was described in seven (70%) (Fig 7a).


Figure 7A
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7a: Graphs illustrate quality criteria for the (a) dynamic CT, (b) dynamic MR, (c) FDG PET, and (d) 99mTc-depreotide SPECT studies assessed in the meta-analysis. The percentages of studies that met the given criteria are shown.

 

Figure 7B
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7b: Graphs illustrate quality criteria for the (a) dynamic CT, (b) dynamic MR, (c) FDG PET, and (d) 99mTc-depreotide SPECT studies assessed in the meta-analysis. The percentages of studies that met the given criteria are shown.

 

Figure 7C
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7c: Graphs illustrate quality criteria for the (a) dynamic CT, (b) dynamic MR, (c) FDG PET, and (d) 99mTc-depreotide SPECT studies assessed in the meta-analysis. The percentages of studies that met the given criteria are shown.

 

Figure 7D
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7d: Graphs illustrate quality criteria for the (a) dynamic CT, (b) dynamic MR, (c) FDG PET, and (d) 99mTc-depreotide SPECT studies assessed in the meta-analysis. The percentages of studies that met the given criteria are shown.

 
With regard to the six dynamic MR trials, study quality scores ranged from 3 to 5. Five (83%) MR trials were performed prospectively, and four (67%) involved 30 or more subjects. Sufficient uniformity of the verification test was reported in one (17%) trial, a sufficient description of the index test was reported in all (100%) six trials, and a sufficient description of the reference-standard test was reported in no trial. Sufficient information on the patients' characteristics was reported in four (67%) trials, but sufficient information on the diagnostic test characteristics was reported in no trial. Patients who broadly reflected the general population were recruited in four (67%) trials, but blinded interpretation of test results was not described in any trial (Fig 7b).

With regard to the 22 FDG PET trials, study quality scores ranged from 4 to 9. Ten (45%) FDG PET trials were performed prospectively, and 17 (77%) involved 30 or more subjects. Sufficient uniformity of the verification test was reported in nine (41%), a sufficient description of the index test was reported in all 22 (100%), and use of the appropriate reference-standard test was reported in 20 (91%) trials. Sufficient information on the patients' characteristics was reported in 17 (77%) trials, and sufficient information on the diagnostic test characteristics was reported in one (4%) trial. Patients who broadly reflected the general population were recruited in 21 (95%) trials, and blinded interpretation of test results was described in 10 (45%) (Fig 7c).

With regard to the seven 99mTc-depreotide SPECT trials, study quality scores ranged from 3 to 6. All seven (100%) SPECT trials were performed prospectively, and five (71%) involved 30 or more subjects. Sufficient uniformity of the verification test was reported in three (43%) trials, a sufficient description of the index test was reported in all seven (100%) trials, but a sufficient description of the reference-standard test was reported in no trials. Sufficient information on the patients' characteristics was reported in four (57%) trials, but sufficient information on the diagnostic test characteristics was reported in no trial. Patients who broadly reflected the general population were recruited in six (86%) trials, and blinded interpretation of test results was described in two (29%) (Fig 7d). Variation in performance measures, with the exception of CT sensitivity, was observed across all modalities. As expected, study quality, as measured by using summary quality scores, accounted for most of the between-study variation, contributing to 95%–97% of the variation.

Heterogeneity
There was high heterogeneity in the study-specific sensitivity for all modalities. Q, P, I2, and degrees of freedom values were, respectively, –698.9, >.99, 100, and 9 for dynamic CT; –167.6, >.99, 100, and 5 for dynamic MR; –539.85, >.99, 100, and 21 for FDG PET; and –293.69, >.99, 100, and 6 for 99mTc-depreotide SPECT. There was also high heterogeneity in the study-specific specificity for all modalities. Q, P, I2, and degrees of freedom values were, respectively, 1043.8, <.01, 99, and 9 for dynamic CT; 510.2, <.01, 99, and 6 for dynamic MR; 1204.6, <.01, 98, and 21 for FDG PET; and 470.5, <.01, 99, and 6 for 99mTc-depreotide SPECT.

Normality Assumption and Publication Bias
The normal quantile plots constructed according to modality (Fig 8) show that there were distinct gaps in the observed data. These gaps suggest the presence of publication bias or separate populations. Nearly all of the data were within the 95% CI bounds; these results suggest that the differences in patient populations or the degree of publication bias was small.


Figure 8
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 8: Normal quantile plots. Gaps in the data indicate different populations and/or publication bias (ie, missing studies). However, the majority of the data points lie between the 95% CI bounds (dashed lines); this indicates that the study estimates are normally distributed. Points outside the 95% CI bounds represent outlier studies that may disproportionately contribute to heterogeneity. DCECT = dynamic contrast-enhanced CT, DCEMRI = dynamic contrast-enhanced MR imaging.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
The evaluation of tumor vascularity with dynamic contrast-enhanced CT has proved to be useful for differentiating malignant from benign nodules. In general, malignant nodules tend to enhance substantially more than benign nodules (26). However, there is some overlap in enhancement between malignant and benign nodules such as active granulomas and benign vascular tumors. Therefore, although the results of these dynamic studies showed high sensitivity for the diagnosis of malignant nodules, the specificity was low (26). In addition, approximately 50% of indeterminate lung nodules, for which the diagnosis is determined at surgery, are benign, and hospitalization for the surgical removal of these nodules is expensive and associated with a certain degree of morbidity and mortality (26).

Several researchers have reported advantages to using dynamic contrast-enhanced MR imaging for tumor characterization. Sequences involving short acquisition times enable one to assess the transit of the first contrast material bolus in tumor perfusion (30). Although the effect of iodinated contrast material in CT examinations is based directly on the concentration of contrast material in the blood, the paramagnetic contrast material effects in MR imaging depend additionally on the interactions of mobile water molecules in all tissue compartments, including the interstitium and cytoplasm; therefore, enhancement measurements are not directly comparable (30). In contrast to MR imaging in other organs, MR imaging of pulmonary nodules is not a standard examination because of known artifacts that result from tissue-air transitions and relatively low spatial resolution. Therefore, there is little information available on dynamic MR imaging of solitary pulmonary nodules (30). However, results of the MR imaging investigations in this meta-analysis have shown sensitivities in the differentiation of malignant and benign solitary pulmonary nodules that are comparable to those obtained with dynamic contrast-enhanced CT, but with higher specificity, especially in the trials of Hittmair et al (27), Guckel et al (28), and Ohno et al (29).

Recent developments show that noninvasive functional imaging may help to distinguish benign from malignant pulmonary nodules. Many reports have proved the value of FDG PET in characterizing the nature of these lesions with use of measurements of their metabolic activity (53). Fluorodeoxyglucose uptake at PET has been evaluated in several ways, such as visual assessment, standard uptake value measurement, and measurement of the tumor uptake–to–normal organ uptake ratio (50). Visual assessment is usually based on comparison of the FDG uptake by the lesion with the uptake in the normal mediastinal blood pool and is the simplest of these analyses. However, nodules with FDG uptake that is similar to the FDG uptake in the mediastinum are difficult to evaluate visually. To assess FDG uptake more objectively, a cutoff standard uptake value of 2.5 has frequently been used (35). However, a number of factors can affect the standard uptake value; these factors include patient body size, blood glucose concentration, time to imaging after injection, and lesion size (50). As a result, the mean standard uptake value for malignant nodules has been reported to range widely from 5.5 to 10.1. More recently, 99mTc-depreotide SPECT with use of a 99mTc-radiolabeled somatostatin analogue has been validated for the diagnosis of malignant solitary pulmonary nodules (53).

In our meta-analysis, we assessed the sensitivity, specificity, PPV, NPV, odds ratio, and area under the ROC curve for four imaging modalities: dynamic CT, dynamic MR imaging, FDG PET, and 99mTc-depreotide SPECT. Although small differences in the point estimates of performance were noted, the 95% CIs excluded significant differences. Our FDG PET results (sensitivity of 0.95, specificity of 0.82) closely resemble those reported by Gould et al (5) in a meta-analysis involving the assessment of 450 solitary pulmonary nodules (93.9% sensitivity, 85.8% specificity). Our results are also in the range of values reported by Bourguet et al (89%–100% sensitivity, 93.5%–100% specificity) (59).

Overall, study methodologic quality was fair. The dynamic MR studies had the worst average quality score, and the dynamic CT studies had the best average score. Furthermore, study quality accounted for 95%–97% of the between-study variation.

To our knowledge, our investigation is the first meta-analysis in which all four modalities—dynamic CT, dynamic MR imaging, FDG PET, and 99mTc-depreotide SPECT—were assessed for the detection of malignancy in solitary pulmonary nodules. To our knowledge, it is also the first meta-analysis in which dynamic CT, dynamic MR imaging, and 99mTc-depreotide SPECT were evaluated individually for the detection of malignancy in solitary pulmonary nodules.

Our study had limitations. First, gaps in the normal quantile plots suggest the presence, albeit small, of publication bias, which may have led to an overestimation of the true diagnostic performance. Different underlying populations in the studies also may have caused gaps in the normal quantile plots. In addition, referral patterns may differ according to imaging technique, and differences in the underlying populations may result in spectrum bias.

Second, we used only published studies in this meta-analysis; therefore, the results are weighted toward the findings of published trials. In meta-analyses, the exclusion of unpublished data is generally associated with an overestimation of the true effect (60). The single most common reason for the inability to publish a trial is the lack of statistical significance; however, for trials of diagnostic tests, the inability to publish may be more dependent on diagnostic accuracy than on statistical significance. Some have suggested that the quality of unpublished data is not comparable to the quality of data reported in articles accepted for publication in peer-reviewed journals (61). Another limitation stemmed from the fact that many of the included studies did not have high quality scores and were, as a whole, heterogeneous. In many studies, whether the investigation met the quality criteria was not specified, so the true study quality was not known.

In conclusion, dynamic contrast-enhanced CT and MR imaging, FDG PET, and 99mTc-depreotide SPECT are accurate noninvasive imaging tests for diagnosing benign versus malignant pulmonary nodules and larger mass lesions. There were negligible differences in performance between the tests evaluated.


   ADVANCE IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
References
 


   FOOTNOTES
 

Abbreviations: CI = confidence interval • FDG = fluorine 18 fluorodeoxyglucose • NPV = negative predictive value • PPV= positive predictive value • ROC = receiver operating characteristic

Author contributions: Guarantors of integrity of entire study, all authors; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, all authors; statistical analysis, all authors; and manuscript editing, all authors

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 

  1. Cronin P. Thoracic CT. Appl Radiol 2003;32:14–27.
  2. Winer-Muram H. The solitary pulmonary nodule. Radiology 2006;239(1):34–49. [Abstract/Free Full Text]
  3. Nolop KB, Rhodes CG, Brudin LH, et al. Glucose utilization in vivo by human pulmonary neoplasms. Cancer 1987;60(11):2682–2689. [CrossRef][Medline]
  4. Worsley DF, Celler A, Adam MJ, et al. Pulmonary nodules: differential diagnosis using 18F-fluorodeoxyglucose single-photon emission computed tomography. AJR Am J Roentgenol 1997;168(3):771–774. [Abstract/Free Full Text]
  5. Gould MK, Maclean CC, Kuschner WG, Rydzak CE, Owens DK. Accuracy of positron emission tomography for diagnosis of pulmonary nodules and mass lesions: a meta-analysis. JAMA 2001;285(7):914–924. [Abstract/Free Full Text]
  6. Blum JE, Handmaker H, Rinne NA. The utility of a somatostatin-type receptor binding peptide radiopharmaceutical (P829) in the evaluation of solitary pulmonary nodules. Chest 1999;115(1):224–232. [CrossRef][Medline]
  7. Awaya H, Matsumoto T, Honjo K, Miura G, Emoto T, Matsunaga N. A preliminary study of discrimination among the components of small pulmonary nodules by MR imaging: correlation between MR images and histologic appearance. Radiat Med 2000;18(1):29–38. [Medline]
  8. Khan KS. Systematic reviews of diagnostic tests: a guide to methods and application. Best Pract Res Clin Obstet Gynaecol 2005;19(1):37–46. [Medline]
  9. Deville WL, Buntinx F, Bouter LM, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Methodol 2002;2:9. [CrossRef][Medline]
  10. Bachmann LM, Bischof DB, Bischofberger SA, Bonani MG, Osann FM, Steurer J. Systematic quantitative overviews of the literature to determine the value of diagnostic tests for predicting acute appendicitis: study protocol. BMC Surg 2002;2:2. [CrossRef][Medline]
  11. Honest H, Khan KS. Reporting of measures of accuracy in systematic reviews of diagnostic literature. BMC Health Serv Res 2002;2(1):4. [CrossRef][Medline]
  12. Coughlin SS, Trock B, Criqui H, Pickle LW, Browner D, Tefft MC. The logistic modeling of sensitivity, specificity and predictive value of a diagnostic test. J Clin Epidemiol 1992;45(1):1–7. [CrossRef][Medline]
  13. Knottnerus JA. Application of logistic regression to the analysis of diagnostic data: exact modeling of a probability tree of multiple binary variables. Med Decis Making 1992;12(2):93–108. [Abstract/Free Full Text]
  14. Greiner M, Gardner IA. Epidemiologic issues in the validation of veterinary diagnostic tests. Prev Vet Med 2000;45(1-2):3–22. [CrossRef][Medline]
  15. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327(7414):557–560. [Free Full Text]
  16. Wang MC, Bushman BJ. Using normal quantile plots to explore meta-analysis data sets. Psychol Methods 1998;3:46–54. [CrossRef]
  17. Swensen SJ, Morin RL, Schueler BA, et al. Solitary pulmonary nodule: CT evaluation of enhancement with iodinated contrast material—a preliminary report. Radiology 1992;182(2):343–347. [Abstract/Free Full Text]
  18. Swensen SJ, Brown LR, Colby TV, Weaver AL. Pulmonary nodules: CT evaluation of enhancement with iodinated contrast material. Radiology 1995;194(2):393–398. [Abstract/Free Full Text]
  19. Yamashita K, Matsunobe S, Tsuda T, et al. Solitary pulmonary nodule: preliminary study of evaluation with incremental dynamic CT. Radiology 1995;194(2):399–405. [Abstract/Free Full Text]
  20. Swensen SJ, Brown LR, Colby TV, Weaver AL, Midthun DE. Lung nodule enhancement at CT: prospective findings. Radiology 1996;201(2):447–455. [Abstract/Free Full Text]
  21. Potente G, Iacari V, Caimi M. The challenge of solitary pulmonary nodules: HRCT evaluation. Comput Med Imaging Graph 1997;21(1):39–46. [CrossRef][Medline]
  22. Zhang M, Kono M. Solitary pulmonary nodules: evaluation of blood flow patterns with dynamic CT. Radiology 1997;205(2):471–478. [Abstract/Free Full Text]
  23. Swensen SJ, Viggiano RW, Midthun DE, et al. Lung nodule enhancement at CT: multicenter study. Radiology 2000;214(1):73–80. [Abstract/Free Full Text]
  24. Yi CA, Lee KS, Kim EA, et al. Solitary pulmonary nodules: dynamic enhanced multi-detector row CT study and comparison with vascular endothelial growth factor and microvessel density. Radiology 2004;233(1):191–199. [Abstract/Free Full Text]
  25. Kim JH, Kim HJ, Lee KH, Kim KH, Lee HL. Solitary pulmonary nodules: a comparative study evaluated with contrast-enhanced dynamic MR imaging and CT. J Comput Assist Tomogr 2004;28(6):766–775. [CrossRef][Medline]
  26. Jeong YJ, Lee KS, Jeong SY, et al. Solitary pulmonary nodule: characterization with combined wash-in and washout features at dynamic multi–detector row CT. Radiology 2005;237(2):675–683. [Abstract/Free Full Text]
  27. Hittmair K, Eckersberger F, Klepetko W, Helbich T, Herold CJ. Evaluation of solitary pulmonary nodules with dynamic contrast-enhanced MR imaging: a promising technique. Magn Reson Imaging 1995;13(7):923–933. [CrossRef][Medline]
  28. Guckel C, Schnabel K, Deimling M, Steinbrich W. Solitary pulmonary nodules: MR evaluation of enhancement patterns with contrast-enhanced dynamic snapshot gradient-echo imaging. Radiology 1996;200(3):681–686. [Abstract/Free Full Text]
  29. Ohno Y, Hatabu H, Takenaka D, Adachi S, Kono M, Sugimura K. Solitary pulmonary nodules: potential role of dynamic MR imaging in management—initial experience. Radiology 2002;224(2):503–511. [Abstract/Free Full Text]
  30. Schaefer JF, Vollmar J, Schick F, et al. Solitary pulmonary nodules: dynamic contrast-enhanced MR imaging—perfusion differences in malignant and benign lesions. Radiology 2004;232(2):544–553. [Abstract/Free Full Text]
  31. Tozaki M, Ichiba N, Fukuda K. Dynamic magnetic resonance imaging of solitary pulmonary nodules: utility of kinetic patterns in differential diagnosis. J Comput Assist Tomogr 2005;29(1):13–19. [CrossRef][Medline]
  32. Kubota K, Matsuzawa T, Fujiwara T, et al. Differential diagnosis of lung tumor with positron emission tomography: a prospective study. J Nucl Med 1990;31(12):1927–1932. [Abstract/Free Full Text]
  33. Gupta NC, Frank AR, Dewan NA, et al. Solitary pulmonary nodules: detection of malignancy with PET with 2-[F18]-fluoro-2-deoxy-D-glucose. Radiology 1992;184(2):441–444. [Abstract/Free Full Text]
  34. Dewan NA, Gupta NC, Redepenning LS, Phalen JJ, Frick MP. Diagnostic efficacy of PET-FDG imaging in solitary pulmonary nodules: potential role in evaluation and management. Chest 1993;104(4):997–1002. [CrossRef][Medline]
  35. Patz EF Jr, Lowe VJ, Hoffman JM, et al. Focal pulmonary abnormalities: evaluation with F-18 fluorodeoxyglucose PET scanning. Radiology 1993;188(2):487–490. [Abstract/Free Full Text]
  36. Dewan NA, Reeb SD, Gupta NC, Gobar LS, Scott WJ. PET-FDG imaging and transthoracic needle lung aspiration biopsy in evaluation of pulmonary lesions: a comparative risk-benefit analysis. Chest 1995;108(2):441–446. [CrossRef][Medline]
  37. Duhaylongsod FG, Lowe VJ, Patz EF Jr, Vaughn AL, Coleman RE, Wolfe WG. Detection of primary and recurrent lung cancer by means of F-18 fluorodeoxyglucose positron emission tomography (FDG PET). J Thorac Cardiovasc Surg 1995;110(1):130–139. [Abstract/Free Full Text]
  38. Duhaylongsod FG, Lowe VJ, Patz EF Jr, 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(5):1348–1352. [Abstract/Free Full Text]
  39. Gupta NC, Maloof J, Gunel E. Probability of malignancy in solitary pulmonary nodules using fluorine-18-FDG and PET. J Nucl Med 1996;37(6):943–948. [Abstract/Free Full Text]
  40. Bury T, Dowlati A, Paulus P, et al. Evaluation of the solitary pulmonary nodule by positron emission tomography imaging. Eur Respir J 1996;9(3):410–414. [Abstract]
  41. Dewan NA, Shehan CJ, Reeb SD, Gobar LS, Scott WJ, Ryschon K. Likelihood of malignancy in a solitary pulmonary nodule: comparison of Bayesian analysis and results of FDG-PET scan. Chest 1997;112(2):416–422. [CrossRef][Medline]
  42. Hagberg RC, Segall GM, Stark P, Burdon TA, Pompili MF. Characterization of pulmonary nodules and mediastinal staging of bronchogenic carcinoma with F-18 fluorodeoxyglucose positron emission tomography. Eur J Cardiothorac Surg 1997;12(1):92–97. [Abstract]
  43. Gupta N, Gill H, Graeber G, Bishop H, Hurst J, Stephens T. Dynamic positron emission tomography with F-18 fluorodeoxyglucose imaging in differentiation of benign from malignant lung/mediastinal lesions. Chest 1998;114(4):1105–1111. [CrossRef][Medline]
  44. Lowe VJ, Fletcher JW, Gobar L, et al. Prospective investigation of positron emission tomography in lung nodules. J Clin Oncol 1998;16(3):1075–1084. [Abstract]
  45. Orino K, Kawamura M, Hatazawa J, Suzuki I, Sazawa Y. Efficacy of F-18 fluorodeoxyglucose positron emission tomography (FDG-PET) scans in diagnosis of pulmonary nodules [in Japanese]. Jpn J Thorac Cardiovasc Surg 1998;46(12):1267–1274. [Medline]
  46. Prauer HW, Weber WA, Romer W, Treumann T, Ziegler SI, Schwaiger M. Controlled prospective study of positron emission tomography using the glucose analogue [18f] fluorodeoxyglucose in the evaluation of pulmonary nodules. Br J Surg 1998;85(11):1506–1511. [CrossRef][Medline]
  47. Lee J, Aronchick JM, Alavi A. Accuracy of F-18 fluorodeoxyglucose positron emission tomography for the evaluation of malignancy in patients presenting with new lung abnormalities: a retrospective review. Chest 2001;120(6):1791–1797. [CrossRef][Medline]
  48. Higashi K, Ueda Y, Sakuma T, et al. Comparison of [(18)F]FDG PET and (201)TI SPECT in evaluation of pulmonary nodules. J Nucl Med 2001;42(10):1489–1496. [Abstract/Free Full Text]
  49. Hung GU, Shiau YC, Tsai SC, Ho YJ, Kao CH, Yen RF. Differentiation of radiographically indeterminate solitary pulmonary nodules with. Jpn J Clin Oncol 2001;31(2):51–54. [Abstract/Free Full Text]
  50. Nomori H, Watanabe K, Ohtsuka T, Naruke T, Suemasu K, Uno K. Visual and semiquantitative analyses for F-18 fluorodeoxyglucose PET scanning in pulmonary nodules 1 cm to 3 cm in size. Ann Thorac Surg 2005;79(3):984–988. [Abstract/Free Full Text]
  51. Herder GJ, Golding RP, Hoekstra OS, et al. The performance of (18)F-fluorodeoxyglucose positron emission tomography in small solitary pulmonary nodules. Eur J Nucl Med Mol Imaging 2004;31(9):1231–1236. [Medline]
  52. Lin WY, Chang GC, Wang SJ. Effectiveness of FDG-PET scans in the evaluation of patients with single pulmonary nodules in Taiwan. Ann Nucl Med Sci 2004;17:63–68.
  53. Halley A, Hugentobler A, Icard P, et al. Efficiency of 18F-FDG and 99mTc-depreotide SPECT in the diagnosis of malignancy of solitary pulmonary nodules. Eur J Nucl Med Mol Imaging 2005;32(9):1026–1032. [CrossRef][Medline]
  54. Blum J, Handmaker H, Lister-James J, Rinne N. A multicenter trial with a somatostatin analog (99m) Tc depreotide in the evaluation of solitary pulmonary nodules. Chest 2000;117(5):1232–1238. [CrossRef][Medline]
  55. Grewal RK, Dadparvar S, Yu JQ, et al. Efficacy of Tc-99m depreotide scintigraphy in the evaluation of solitary pulmonary nodules. Cancer J 2002;8(5):400–404. [Medline]
  56. Baath M, Kolbeck KG, Danielsson R. Somatostatin receptor scintigraphy with 99mTc-depreotide (NeoSpect) in discriminating between malignant and benign lesions in the diagnosis of lung cancer: a pilot study. Acta Radiol 2004;45(8):833–839. [CrossRef][Medline]
  57. Kahn D, Menda Y, Kernstine K, et al. The utility of 99mTc depreotide compared with F-18 fluorodeoxyglucose positron emission tomography and surgical staging in patients with suspected non-small cell lung cancer. Chest 2004;125(2):494–501. [CrossRef][Medline]
  58. Martins T, Lino JS, Ramos S, Oliveira L. 99mTc-depreotide scintigraphy in the evaluation of indeterminate pulmonary lesions: clinical experience. Cancer Biother Radiopharm 2004;19(2):253–259. [CrossRef][Medline]
  59. Bourguet P; for Groupe de Travail SOR. 2002 standards, options and recommendations for the use of [18F]-FDG (PET-FDG) in cancerology [in French]. Bull Cancer 2003;90(Spec Issue):S1–S109. [Medline]
  60. Trivedi HS, Moore H, Nasr S, et al. A randomized prospective trial to assess the role of saline hydration on the development of contrast nephrotoxicity. Nephron Clin Pract 2003;93(1):C29–C34. [CrossRef][Medline]
  61. McAuley L, Pham B, Tugwell P, Moher D. Does the inclusion of grey literature influence estimates of intervention effectiveness reported in meta-analyses? Lancet 2000;356(9237):1228–1231. [CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supplemental Tables
Right arrow All Versions of this Article:
2463062148v1
246/3/772    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cronin, P.
Right arrow Articles by Carlos, R. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cronin, P.
Right arrow Articles by Carlos, R. C.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE