SYMPOSIA
Measurement Methods and Algorithms for the Management of Solid Nodules Arjun Nair, MBChB, MRCP, FRCR,* David R. Baldwin, MD, FRCP,w John K. Field, PhD, FRCPath,z David M. Hansell, MD, FRCP, FRCR,* and Anand Devaraj, MD, MRCP, FRCRy
Abstract: The increased detection of incidental small pulmonary nodules on multidetector computed tomography has driven attempts to refine the characterization and management of such nodules. A variety of methods have been developed to measure the size and biological activity of nodules to help define their nature, but these have limitations. Several clinical trials have assessed the efficacy of low-dose computed tomography screening for lung cancer and offer some insights into these limitations; however, they also provide evidence that refines existing nodule management strategies. This article reviews the size-based and functional measurement methods that can be used to predict the likelihood of malignancy in noncalcified solid pulmonary nodules and discusses their incorporation into existing algorithms for nodule management. The issue of multiple nodules and the optimum frequency and duration of follow-up are explored. Key Words: pulmonary nodules, low-dose computed tomography screening, volumetry, Fleischner guidelines
trials have also shown that the vast majority of solid nodules <10 mm are benign, even in high-risk populations. Older screening trials have been instrumental in informing guidelines for radiologic follow-up of nodules.2 In the near future, data from more recent and ongoing screening trials, such as the National Lung Screening Trial (NLST) in the United States (the largest randomized trial ever conducted comparing LDCT with chest radiography),5 should help to refine these nodule management strategies. Any nodule management guideline should have, as a minimum, guidance on the measurement of nodule size and growth. This article reviews current knowledge regarding measurement methodology and discusses its incorporation into solid nodule management strategies. Particular reference is made to observations extrapolated from the LDCT screening trials. Part-solid nodules are discussed in another article in this issue.
(J Thorac Imaging 2012;27:230–239)
CONSIDERATIONS AT THE OUTSET
T
he detection of incidental pulmonary nodules has become commonplace as a consequence of the widespread availability of multidetector computed tomography (MDCT).1 The distinction of benign from malignant nodules has thus become a frequent dilemma. There is now a large body of literature providing guidance on the detection, characterization, and management of such nodules.2–4 Such guidance assimilates various clinical factors in conjunction with sizebased, morphologic, and occasionally metabolic characteristics to assess the likelihood of malignancy in nodules. However, once such an assessment has been carried out, many noncalcified nodules (whether solid, part-solid, or nonsolid) will still be indeterminate and will require follow-up CT. These indeterminate nodules are a constant challenge to both radiologists and clinicians. The results of multiple trials that have assessed screening for lung cancer with low-dose computed tomography (LDCT) have underscored the ability of CT to detect lung cancer at a potentially curable stage.5–12 However, these From the *Department of Radiology, Royal Brompton Hospital; yDepartment of Radiology, St George’s Hospital, London; wRespiratory Medicine Unit, David Evans Research Centre, Nottingham University Hospitals, Nottingham; and zRoy Castle Lung Cancer Research Programme, The University of Liverpool Cancer Research Centre, Institute of Translational Medicine, Liverpool, UK. The authors declare no conflicts of interest. Reprints: Arjun Nair, MBChB, MRCP, FRCR, Department of Radiology, Royal Brompton Hospital, Sydney Street, London SW3 6NP, UK (e-mail: arjun7764@gmail.com). Copyright r 2012 by Lippincott Williams & Wilkins
230 | www.thoracicimaging.com
Pulmonary Nodule Definition Determining whether an opacity can be designated a pulmonary nodule on CT may not be straightforward. The Fleischner Society glossary defines a solid pulmonary nodule as a rounded or irregular opacity of homogenous soft tissue attenuation, well or poorly defined, measuring up to 3 cm in diameter.13 It reserves the term micronodule for opacities measuring <3 mm in diameter. The vast majority of opacities designated as solid nodules probably conform to these definitions. However, the increased spatial resolution of MDCT has also resulted in the visualization of opacities that lack easily definable 3-dimensional (3D) shapes or attenuation and so evade precise morphologic definition (Fig. 1). The true prevalence, nature, and propensity for malignant behavior of such abnormalities have not been fully explored yet, but they can be an important source of false-positive nodule designation in both screening and clinical practice.
Solitary or Multiple Nodules Historically, the management of incidental nodules has focused on the solitary pulmonary nodule. However, multiple pulmonary nodules are a frequent finding on CT, and this has been the case in screening studies. Of those subjects with at least 1 noncalcified nodule, 31.8% in the Early Lung Cancer Action Project (ELCAP)14 and 29.3% in the Italian Lung Cancer Computed Tomography Screening Trial (ITALUNG)8 had multiple noncalcified nodules. Current recommendations and screening algorithms do not, however, specifically distinguish between single and multiple nodules. In the absence of known J Thorac Imaging
Volume 27, Number 4, July 2012
J Thorac Imaging
Volume 27, Number 4, July 2012
Management of Pulmonary Nodules
MEASUREMENT METHODS FOR SOLID PULMONARY NODULES Both the size and growth rate of a pulmonary nodule correlate strongly with its likelihood of malignancy.3 The measurement of size is thus crucial for 2 reasons: (1) to stratify risk at baseline CT; and (2) to detect an objective change that may signify growth and probability of malignancy at follow-up CT. Any nodule measurement method that is to be useful in clinical practice must thus be able to maximize the chance of detecting cancer at an early stage while minimizing the requirement to perform unnecessary invasive tests. At the same time, measurement methods need to demonstrate adequate accuracy, reproducibility, ease of use, and convenience of integration into the workflow. Assessment of the size and growth of solid nodules can be performed using diameter measurements or volumetric estimation. Currently, manual diameter estimation using electronic calipers (whether as a single maximum diameter or the average of short-axis and long-axis measurements as advocated by the Fleischner society recommendations2) is the most widely used parameter in clinical practice. This is because it is quick and convenient to perform. Nodule volume, in contrast, can be either manually derived, which relies on user-delineated nodule boundaries on axial sections that are then translated into a 3D volume by software, or semiautomated, using segmentation algorithms.15 Semiautomated volumetry is also quick and simple to perform, requiring only a mouse click to initiate the process. However, semiautomated volumetry is not always a readily accessible tool, as it usually requires a dedicated workstation rather than integration into routine picture archive and communication workflow.
Reliability of Diameter Measurements
FIGURE 1. LDCT in a 63-year-old male ex-smoker demonstrates a 4 mm opacity (arrow) in the right upper lobe on an axial 1 mm collimation image on lung windows (A), more clearly seen on the maximum intensity projection image (B). However, this opacity appears round on only 1 thin-section image, and in other planes appears flat or comma-shaped, as demonstrated by a 3D volumetric segmentation (C). It is arguable whether such opacities should be designated nodules.
patterns of infection or inflammation, most radiologists would likely choose a management strategy based on the largest single nodule identified. However, in the setting of a multiplicity of nodules, deciding how many nodules should be marked and targeted for surveillance and for how long can be difficult. r
2012 Lippincott Williams & Wilkins
The measurement of diameter with electronic calipers is subject to considerable intrareader and interreader variability16–18 (Fig. 2). This was demonstrated in a nonscreening cohort by Revel and colleagues. They showed that, when measuring nodules 2 cm or less, the limits of intraobserver and interobserver variability were 1.32 and 1.73 mm, respectively. This means that a nodule could confidently be said to have grown only if its diameter had increased beyond these limits.18 For example, because only a 26% increase in diameter of a spherical nodule is required to represent a single volume doubling,19 it could be falsely concluded that a nodule measuring 5 mm at baseline and then 6.3 mm at 3-month follow-up has doubled in volume when it is in fact stable and the difference is due to measurement error. In contrast, measurement error may result in growing nodules that are classified as stable.
Reliability of Volume Measurements Volume measurements can be expected to be more accurate than unidimensional or bidimensional measurements alone, as a volumetric measurement evaluates the entirety of a nodule.20–22 In a study of 322 synthetic nodules implanted in porcine lungs, Bolte et al23 demonstrated that volumetry could estimate true lesion size to within a mean deviation of 9.2% for semiautomated and 0.3% for manual-corrected volumetry. Volumetric software-derived measurements also demonstrate good reproducibility9,22,24 and high interobserver agreement.9,25,26 The measurement error between volumetric measurements of the same nodule is up to about 27%, with the majority of volumetric www.thoracicimaging.com |
231
J Thorac Imaging
Nair et al
Volume 27, Number 4, July 2012
FIGURE 2. The maximum diameter of this nodule in the apical segment of the left lower lobe has apparently increased from 9.3 mm (A) to 10.9 mm (B) on 2 CT scans performed 6 months apart. This 1.6 mm difference may represent measurement error between readers when 2D measurements using electronic calipers are used.
measurements demonstrating variability of <10%.9,25 Multiple factors may be responsible for such variation; some of the key factors are summarized in Table 1 (Fig. 3).
Volume Doubling Time Limitations in the accuracy and reproducibility of diameter measurements have led to the use of volumetric
TABLE 1. Selected Studies Investigating Various Parameters That Affect Reproducibility and Accuracy of Volumetric Measurement
References
Type/Source of Nodules
Parameter(s) Assessed
Ashraf et al27
In vivo, screening subjects
Three segmentation algorithms for nodules (all sizes, small sized, and subsolid)
Christe et al28
In vivo, screening subjects and cancer patients
Tube current time
Gietema et al25
In vivo, screening subjects
Interobserver variability
Gietema et al29
In vivo, screening subjects
Nodule size, segmentation performance, inspiration level
Goo et al30
Ex vivo, synthetic
Threshold density for segmentation, section thickness, FOV, reconstruction interval
Ko et al31
Ex vivo, synthetic
Tube current time, reconstruction algorithm, nodule size, nodule attenuation, method of volumetric measurement
Ravenel et al32
Ex vivo, synthetic
Reconstruction algorithm, FOV, section thickness
Wang et al33
In vivo, screening subjects
Nodule size, morphology, location
Wang et al34
In vivo, screening subjects
Reconstruction algorithm, slice thickness
232 | www.thoracicimaging.com
Key Conclusions Choosing different segmentation algorithms within the same software package led to increased interobserver variability Automated volumetric measurements at 10 mAs, comparable with that at 300 mAs, suggest that 10 mAs could be used as the lowest tube current-time threshold for nodule follow-up Causes for discrepant readings were incomplete segmentation in nodules <30 mm3, and irregular margins in larger nodules Segmentation accuracy heavily dependent on: nodule shape (incomplete segmentation when nonspherical) margin (incomplete segmentation when irregular) Inspiratory level only has a weak effect on perfectly segmented nodules Decreased measurement error associated with increasing nodule diameter decreasing section thickness segmentation thresholds of 400 and 500 HU Decreased measurement error associated with high-frequency (hard) reconstruction algorithms (for nodules <5 mm) increasing nodule diameter higher tube current time (120 mAs) No substantial effect of location of nodule on accuracy Decreased measurement error associated with increasing nodule diameter decreasing section thickness No substantial difference in accuracy across reconstruction algorithms, or different FOVs Segmentation variability is high with irregular, juxtavascular nodules (Fig. 3) moderate with spiculated, perifissural or pleuralbased nodules low with lobulated or larger (50-500 mm3) nodules 1-mm-thick slices with low frequency (soft) reconstruction algorithm provides the most repeatability
r
2012 Lippincott Williams & Wilkins
J Thorac Imaging
Volume 27, Number 4, July 2012
Management of Pulmonary Nodules
FIGURE 3. A juxtapleural nodule (A) is not accurately segmented by semiautomated 3D volumetric segmentation (B) using a commercially available software package (Lungcare, Siemens Medical Solutions, Erlangen, Germany) because of difficulties defining its contour and surface where it is adjacent to the pleura. The yellow region represents the segmented volume.
measurement as the main measure of size and growth in some of the more recent lung screening studies.10,35,36 In a study that compared the reliability of diameter and volume measurements in patients with lung cancers <3 cm in size, growth was found to be consistently inaccurately estimated (and usually underestimated) using diameter and area measurements as compared with volumetric measurements.20 The most commonly used parameter for estimating volumetric growth is volume doubling time (VDT), which presumes an exponential pattern of growth. The VDT of malignant nodules is highly variable but generally lies within the range of 20 and 400 days,22,37–41 with shorter VDTs noted for small cell lung cancers. However, between 5%42 and 27%43 of malignant nodules may have VDTs > 400 days, which has led some to question whether these cases represent overdiagnosis—that is, the diagnosis of cancers that are so indolent or slow growing that an individual is likely to die from some other cause before the cancer becomes apparent.43 In any event, a purely exponential growth model may be too simplistic, as illustrated by Lindell et al38; they found that growth curves of 18 cancers could indeed remain flat or demonstrate sudden accelerated growth. More recently, Ko and colleagues assessed the utility of an alternative “normative” model for growth rate determination in a screening population, which accounts for potentially coexistent patterns of growth (eg, linear, exponential, and Gompertzian growth). They found that this experimental model had the potential to successfully distinguish between benign and malignant nodules, using volumetric 3D assessment.44
MANAGEMENT ALGORITHMS FOR INDETERMINATE SOLID PULMONARY NODULES Any nodule management algorithm should maximize the chances of lung cancer detection, minimize false-positive rates, avoid unnecessary anxiety to the patient, and avert undue harm as a result of repeated radiation exposure or unnecessary invasive procedures, while also being cost-effective. The Fleischner guidelines for small solid pulmonary nodule management is one of several algorithms and is well known to both thoracic and nonthoracic radiologists. For the most part, the various nodule management protocols r
2012 Lippincott Williams & Wilkins
adopted by screening studies in the post-Fleischner guidelines era have reinforced the basic tenets of the Fleischner guidelines but with some modifications.
The Fleischner Society Guidelines In 2005, the Fleischner Society proposed guidelines for the management of solid noncalcified pulmonary nodules <8 mm in order to help rationalize the follow-up of the increase in the detection of small nodules largely created by the availability of MDCT.2 These guidelines are predominantly informed by data from screening studies such as ELCAP and by data regarding nodule growth rates from nonscreening experimental studies. The guidelines are only applicable to patients who are aged 35 years or older, with no history of known malignancy. They introduced the concept of follow-up intervals based on mean nodule diameter, and a binary high-risk or low-risk stratification for malignancy based on pretest probability (Table 2). As size increases, shorter periods of follow-up are recommended, but stability of nodules allows increasing time intervals between follow-up scans. According to the Fleischner guidelines, a nodule that is stable over 2 years should not require any further follow-up.
Limitations of the Fleischner Guidelines The Fleischner guidelines undoubtedly continue to provide guidance to radiologists faced with an indeterminate pulmonary nodule. However, they are not without limitations. First, like all screening-derived recommendations, the guidelines are by definition more applicable to a highrisk population, with a consequently potentially more aggressive approach toward abnormalities that could signify malignancy. Further, the guidelines do not address the distinction between solitary and multiple nodules. The sizebased risk-stratified approach of the Fleischner guidelines was a departure from the notion that every nodule, regardless of size, needed a minimum follow-up period of 2 years, which had prevailed before these guidelines. However, the lack of guidance for multiple nodules means that the Fleischner guidelines have the potential to subject the patient to prolonged cycles of CT follow-up if new nodules are detected during the surveillance of an incident nodule www.thoracicimaging.com |
233
234 | www.thoracicimaging.com
50-500
5
r
CT at 6-12 mo, then 18-24 mo if unchanged
CT at 12 mo; if unchanged, no further follow-up
LDCT at 3, 6, 12, or 24 mo, depending on lesion size and level of suspicion of malignancy
None
NLST5
LDCT at 12 mo
None
UKLS36*
Referral to multidisciplinary team for workup
LDCT at 3 mo; LDCT at 3 mo; assessment of assessment of VDT; if VDT; if VDT < 400 VDT < 400 days—refer to days—refer to pulmonologist pulmonologist VDT > 400 VDT = 400-600 days—LDCT days—LDCT at 9 mo at 9 mo VDT > 600 days— LDCT at 9 mo
LDCT at 12 mo
NELSON11
Follow-up Recommendations
CT at 6-12 mo, then CT at 3-6 mo, then 918-24 mo if 12 mo, and 24 mo if unchanged unchanged CT at 3, 9, and 24 mo CT at 3, 9, and 24 mo if unchanged, FDGif unchanged, FDGPET, dynamic PET, dynamic Referral to LDCT at 3, 6, 12, or 24 mo, contrast-enhanced contrast-enhanced pulmonologist depending on lesion size and level CT, and/or biopsy CT, and/or biopsy for workup of suspicion of malignancy; FDG-PET, dynamic contrastenhanced CT, and/or biopsy
CT at 12 mo; if unchanged, no further follow-up
High Risk
No follow-up
Low Risk
Fleischner Guidelines
20052
Combination of VDT assessment and FDGPET for nodules 5-15 mm
DLCST10,45
Nonsmooth
Discretionary oral antibiotics, then LDCT or standard contrastenhanced CT, and FDG-PET
Oral antibiotics, LDCT in 6-8 wk; if no regression, FDG-PET
Oral antibiotics, LDCT in 6-8 wk; if no regression, follow-up CT or invasive procedure
LDCT at 3, 6, LDCT at 3, 6, and 12 mo; if and 12 mo; unchanged, if further CT at unchanged, 24 mo further CT at 24 mo
Smooth
DANTE7
J Thorac Imaging
*Nodules meeting all criteria for intrapulmonary lymph nodes are considered benign in UKLS. DANTE indicates Detection And screening of early lung cancer by Novel imaging Technology and molecular Essays; DLCST, Danish Lung Cancer Screening Trial; NELSON, Nederlands-Leuvens Longkanker Screenings Onderzoek; UKLS, United Kingdom Lung Screening.
> 20
9 10 15
8
7
> 500
15-50
3 4
6
< 15
Volume (mm3)
<2
Diameter (mm)
Noncalcified Nodules
TABLE 2. Comparison of Follow-up Strategies From Selected Lung Screening Trials and the 2005 Fleischner Guidelines
Nair et al Volume 27, Number 4, July 2012
2012 Lippincott Williams & Wilkins
J Thorac Imaging
Volume 27, Number 4, July 2012
over the 2-year period. The validity of stability over a 2-year period as a marker of benignity has also been challenged, albeit more so for part-solid nodules with slow growth.44,46 A recent survey of members of the Society of Thoracic Radiology using 13 clinical scenarios by Esmaili and colleagues highlighted the variation in compliance with the Fleischner guidelines 6 years after their publication. Of the 181 respondents, only 27% made the appropriate recommendation according to the guidelines, and there was a general trend toward “overmanagement.”47 Such a trend indicates the difficulty (whether real or perceived) that radiologists may face in ceasing investigation of a stable abnormality despite the support of evidence-based guidelines.
Potential Implications of Recent Screening Trials for Nodule Management Algorithms The rates of positive tests, lung cancer detection, and invasive procedures and the predominant nodule management strategies used in a selection of recent trials of lung screening are summarized in Tables 2 and 3, respectively. In general, there is little variation in lung cancer detection rates and invasive procedures among current trials, and as yet there is no particular management strategy that is clearly superior. Regardless of the chosen algorithm, the screening trials remind us that LDCT protocols can, and so wherever possible should, be used in the follow-up of incidental lung nodules.
Algorithms Based on Size and Growth The use of diameter was the chosen method of growth assessment in trials such as the NLST and ITALUNG.5,8 As has been discussed, the inherent variability of diameter measurements that has been demonstrated in experimental studies may cause the overestimation or underestimation of nodule growth. However, by extrapolating evidence from the ITALUNG trial, it can be argued that overestimation of growth may not be as much of an issue in clinical practice. Despite using a low threshold of just a 1 mm diameter difference to signify growth in the ITALUNG trial, only 5 of the 366 nodules measuring between 5 and 7 mm that had required follow-up demonstrated growth using this definition, and all of these proved to be lung cancer.8 The results may be explained by the tendency of a radiologist to also use clinical judgment in the evaluation of growth rather than rely on electronic calipers alone. For example, he or she may use subjective factors such as relative proximity or size to other structures (eg, vessels, bronchi, or pleura) to assess the nodule before defining it as growing. The ability of radiologists to “eye-ball” nodules side by side when determining growth is not something that has been easily captured in experimental studies so far. The optimum interval between follow-up scans and the total duration of follow-up to determine nodule stability has not been clearly established. The Fleischner recommendations suggest that nodules between 4 and 8 mm should be followed up for a period of 2 years.2 However, the ability of volumetric analysis to more accurately define nodule stability may allow for a shorter duration of followup. In the NELSON trial, for example, nodules identified at baseline were followed up for a maximum of 12 months.35 The NELSON trial has also suggested a role for volumetry in decreasing the number of CT scans in a follow-up algorithm, specifically for newly identified nodules.35 The r
2012 Lippincott Williams & Wilkins
Management of Pulmonary Nodules
NELSON algorithm allowed the classification of nodules with volumes between 50 and 500 mm3 (corresponding approximately to nodule diameters of between 4.6 and 9.8 mm) as indeterminate. Such subjects were recalled for a repeat CT in 3 months during baseline screening, and this interval was reduced to just 6 to 8 weeks in the incidence rounds. If at 6 to 8 weeks, the VDT was >400 days, the subject was followed up with a further CT at a minimum of 1 year. The rationale for a shorter initial interval of followup CT during incidence screening is that any new malignant nodule that had appeared over the 1-year period between baseline and first incidence screen must have a short VDT and therefore would be shown to be growing rapidly even after a brief period of follow-up. Thus, in the clinical situation in which a new nodule is identified in a patient with a recently performed normal CT, an initial rapid follow-up CT could be recommended, followed, if necessary, by a single CT at 1 year or by early biopsy. Ultimately, however, a nodule management strategy can be considered successful only if it is able to draw the line under further investigation to the satisfaction of the patient, clinician, and radiologist. For this reason, there may always be a tendency for longer and more frequent followup in any nodule management algorithm.
Algorithms Based on Morphologic and Location Characteristics Information from current screening trials has reinforced the idea that, although some morphologic characteristics of solid nodules, such as a smooth or lobulated border, may provide clues to a benign or malignant etiology,48 they are not sufficiently discriminatory and are not as useful as the assessment of size and growth. In a subanalysis of the NELSON study, Xu et al49 found that there was a correlation between lung cancer and a lobulated margin or an irregular shape; however, on multivariate analysis these factors demonstrated only weak correlations, thus limiting their diagnostic utility. In addition, NELSON investigators found no evidence of malignancy in nodules that had “attachment” to the pleura, vessels, or fissures.50 The lack of malignancy in perifissural nodules was also found by Ahn et al.51 In this LDCT screening study, the majority of perifissural nodules were triangular or ovoid (44% and 42%, respectively) and had a septal connection (73%). This led the authors to postulate that the majority of these nodules were intrapulmonary lymph nodes (IPLNs). In recent years, IPLNs have been well recognized as a source of multiple pulmonary nodules, with some histopathology-corroborated CT features described in studies with relatively small samples.52–55 Typically, IPLNs are up to 12 mm in maximum diameter, are polygonal or coffee-bean shaped, lie within 15 mm of a pleural surface, have a smooth surface, and have at least 1 linear opacity connecting it to the pleural surface, reflecting an interlobular septum53,54 (Fig. 4). They are relatively uncommon in the upper lobes. Among the screening trials, the recently launched UK Lung Screening trial recognizes IPLNs as benign entities within its prescribed management algorithm, although it uses a more stringent set of criteria for their diagnosis than those described above. These are that IPLNS should measure <8 mm, lie within 5 mm of a pleural surface or within an interlobar fissure, and meet the other morphologic criteria described above. www.thoracicimaging.com |
235
236 | www.thoracicimaging.com NELSON11,64
DLCST10
3.8
2.9
1 53,454 26,309 27.3 270 1.0 24,102 16.8 211 0.9
24,715 27.9 168 0.7 63.0z 1.8 29.8z 2.4 5.2
2.1
3
2
1 15,822 7557 20.8 70 0.9 63.9 1.2 27.2 4.6 7289 7.8 54 0.5 73.7 0.8 21.3 9.5
2
1 4104 2052 8.7 17 0.8 53 1.2 32 9.5
Mean diameter Volume and diameter Volume and diameter Devolved to local practice; positive Nodules classified as positive, Nodules 5-15 mm classified finding = NCN Z4 mm or any indeterminate, or negative; indeterminate as indeterminate other finding suspicious for lung nodules; indeterminate nodules rescanned in cancer 3 mo (first round) and 6-8 wk (second (eg, mediastinal lymphadenopathy) round) to estimate percentage volume change
NLST5
1 2472 1276 15.6 28 2.2 57 4.2 18.9 NR
1276 27.5 60 4.7 55 7.5 NR NR
3
Diameter Antibiotics, rescanning and/or enhanced CT or PET scan depending on lesion size and suspicion
DANTE7,65
1 3206 1406 30.3 21 1.5 47.6 2.3 NR 4.7
Diameter Positive finding: NCNZ5 mm
ITALUNG8
Only nodule management relevant to solid noncalcified nodules is mentioned. *Positivity rate is defined as the number of subjects with a positive finding, divided by the total number of subjects screened in the LDCT arm, expressed as a percentage. Note that the nodule management strategy of the NELSON study allowed studies to be called indeterminate and subjected to follow-up scans, but for the purposes of standardization with other trials, indeterminate scans have been considered as positive and included in the calculation of the positivity rate. wThe ways in which invasive procedures have been defined and reported so far has varied between trials. For example, the rates of invasive procedures for NLST and DANTE shown here include bronchoscopy, while those of NELSON do not. zThese results reflect the cumulative rates of stage I lung cancer, and percentage of patients who underwent invasive procedures in whom lung cancer was not confirmed, respectively. yPositive predictive value has been calculated as the number of subjects with lung cancer, divided by the total number of subjects with a positive finding in the LDCT arm. DANTE indicates Detection And screening of early lung cancer by Novel imaging Technology and molecular Essays; DLCST, Danish Lung Cancer Screening Trial; NCN, noncalcified nodule; NELSON, Nederlands-Leuvens Longkanker Screenings Onderzoek; NR, not reported.
Screening round No. recruited No. screened in LDCT arm Positivity rate *(%) No. lung cancers in LDCT arm Lung cancer detection rate (%) Stage I cancer (%) Invasive procedures w(%) % with no lung cancer Positive predictive value y(%)
Parameter assessed Nodule management strategy
Trials
TABLE 3. Screening Strategies and Results From Recent Randomized Lung Cancer Screening Trials
Nair et al J Thorac Imaging
r
Volume 27, Number 4, July 2012
2012 Lippincott Williams & Wilkins
J Thorac Imaging
Volume 27, Number 4, July 2012
Management of Pulmonary Nodules
FIGURE 4. Five-millimeter triangular nodule in the middle lobe (arrow) on an axial 1 mm collimation CT image (A) is seen to lie on the minor fissure on a sagittal reconstruction (B). The shape, size, and location of such a nodule is highly suggestive of an intrapulmonary lymph node.
Algorithms Based on Functional Characteristics [18F]-Fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET) and integrated PET-CT have well-defined roles in the staging of non–small cell lung cancer.56 FDG-PET can demonstrate increased metabolic activity indicating malignancy in noncalcified nodules >10 mm in diameter with a sensitivity of 96.8% and specificity of 77.8%.57 However, its use in differentiating benign from malignant small nodules is more questionable, because the sensitivity of FDG-PET is low in nodules <10 mm. This has led to recommendations by the American College of Chest Physicians that nodules measuring at least 8 to 10 mm with a low to moderate pretest probability for malignancy should be referred for FDG-PET for further characterization; there is certainly little value to be gained in FDG-PET workup for smaller nodules.3 FDG-PET is also costly58 and delivers a higher radiation dose59 compared with standard or LDCT. The role of integrated PET-CT in nodule characterization is yet to be extensively evaluated,60 and in this regard the 2 Italian screening studies [ITALUNG and the Continuous Observation of Smoking Subjects (COSMOS) study] incorporating PET and PET-CT into their algorithms for nodule management may throw some light on this matter. Both these studies have used PET and PET-CT, respectively, to characterize solid noncalcified nodules >8 mm. Veronesi et al12 found that the diagnostic sensitivity of PET-CT in such cases was 88%, but it increased to 100% for solid noncalcified nodules >10 mm, in an analysis of the COSMOS study. In a later analysis they have suggested that decreasing the maximum standard uptake value to 1.5 from 2.0 improves sensitivity without compromising specificity in noncalcified nodules <10 mm, but this will require further validation.61 The amalgamation of size-based and functional metrics to enable better characterization of nodules is another attractive prospect in refining nodule management. In a recent analysis from the Danish Lung Cancer Screening Trial, Ashraf et al45 have demonstrated how, in indeterminate nodules (defined as 5 to 15 mm in maximal diameter according to the DLCST protocol), a combination of VDT assessment at 3 months and FDG-PET is superior to either metric alone in predicting malignancy. These studies pror
2012 Lippincott Williams & Wilkins
vide important evidence for the role of FDG-PET in nodule management, but, although FDG-PET may have a wider role in future nodule management algorithms, for the moment it is probably more reliably used for characterizing larger nodules. Progressive enhancement of nodules on dynamic contrast-enhanced CT is another potential method of assessing the functional characteristics of a nodule. This technique uses measurable contrast enhancement above a defined threshold as a surrogate for the increased vascularity within malignant nodules. The threshold for classifying enhancement as malignant can be varied between 15 and 30 HU with corresponding sensitivities and specificities of 98% and 58% and 99% and 54%, respectively.60,62 This method of assessing biological activity has shown good correlation with angiogenesis, as measured by vascular endothelial growth factor of nodules on subsequent immunohistochemical staining of pathologic samples,63 but has not found wide acceptance. In the setting of screening tests, it has only been used thus far in the COSMOS study, in which a subset of 54 subjects with intraparenchymal (not perihilar) nodules measuring >8 mm underwent dynamic contrast-enhanced CT, resulting in 100% sensitivity but low 59% specificity for lung cancer.61
CONCLUSIONS The management of incidental solid pulmonary nodules is an ongoing and evolving radiologic challenge fueled by the greater detection of single or multiple small nodules resulting from increased use of CT. The evaluation of solid noncalcified nodules still relies mainly on size but only as part of a comprehensive risk assessment to identify the pretest probability of malignancy in that particular individual. The proper path for follow-up and diagnosis can be pursued once this risk stratification has been performed. The multiple LDCT lung screening programs are slowly providing information that can be amalgamated with current concepts and algorithms for nodule management to increase their robustness, such that the correct balance is struck between detection of malignancy and avoidance of over investigation. www.thoracicimaging.com |
237
J Thorac Imaging
Nair et al
REFERENCES 1. Fischbach F, Knollmann F, Griesshaber V, et al. Detection of pulmonary nodules by multislice computed tomography: improved detection rate with reduced slice thickness. Eur Radiol. 2003;13:2378–2383. 2. MacMahon H, Austin JH, Gamsu G, et al. Guidelines for management of small pulmonary nodules detected on CT scans: a statement from the Fleischner Society. Radiology. 2005;237:395–400. 3. Gould MK, Fletcher J, Iannettoni MD, et al. Evaluation of patients with pulmonary nodules: when is it lung cancer?: ACCP evidence-based clinical practice guidelines (2nd edition). Chest. 2007;132(3, suppl):108S–130S. 4. Wahidi MM, Govert JA, Goudar RK, et al. Evidence for the treatment of patients with pulmonary nodules: when is it lung cancer? ACCP evidence-based clinical practice guidelines (2nd edition). Chest. 2007;132(3, suppl):94S–107S. 5. Aberle DR, Adams AM, Berg CD, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395–409. 6. Diederich S, Wormanns D, Semik M, et al. Screening for early lung cancer with low-dose spiral CT: prevalence in 817 asymptomatic smokers. Radiology. 2002;222:773–781. 7. Infante M, Lutman FR, Cavuto S, et al. Lung cancer screening with spiral CT: baseline results of the randomized DANTE trial. Lung Cancer. 2008;59:355–363. 8. Lopes PA, Picozzi G, Mascalchi M, et al. Design, recruitment and baseline results of the ITALUNG trial for lung cancer screening with low-dose CT. Lung Cancer. 2009;64: 34–40. 9. Marchiano A, Calabro E, Civelli E, et al. Pulmonary nodules: volume repeatability at multidetector CT lung cancer screening. Radiology. 2009;251:919–925. 10. Pedersen JH, Ashraf H, Dirksen A, et al. The Danish randomized lung cancer CT screening trial—overall design and results of the prevalence round. J Thorac Oncol. 2009;4: 608–614. 11. van Klaveren RJ, Oudkerk M, Prokop M, et al. Management of lung nodules detected by volume CT scanning. N Engl J Med. 2009;361:2221–2229. 12. Veronesi G, Bellomi M, Veronesi U, et al. Role of positron emission tomography scanning in the management of lung nodules detected at baseline computed tomography screening. Ann Thorac Surg. 2007;84:959–965. 13. Hansell DM, Bankier AA, MacMahon H, et al. Fleischner Society: glossary of terms for thoracic imaging. Radiology. 2008;246:697–722. 14. Henschke CI, McCauley DI, Yankelevitz DF, et al. Early Lung Cancer Action Project: overall design and findings from baseline screening. Lancet. 1999;354:99–105. 15. Gavrielides MA, Kinnard LM, Myers KJ, et al. Noncalcified lung nodules: volumetric assessment with thoracic CT. Radiology. 2009;251:26–37. 16. Bogot NR, Kazerooni EA, Kelly AM, et al. Interobserver and intraobserver variability in the assessment of pulmonary nodule size on CT using film and computer display methods. Acad Radiol. 2005;12:948–956. 17. Reeves AP, Biancardi AM, Apanasovich TV, et al. The Lung Image Database Consortium (LIDC): a comparison of different size metrics for pulmonary nodule measurements. Acad Radiol. 2007;14:1475–1485. 18. Revel MP, Bissery A, Bienvenu M, et al. Are two-dimensional CT measurements of small noncalcified pulmonary nodules reliable? Radiology. 2004;231:453–458. 19. Lillington GA, Caskey CI. Evaluation and management of solitary and multiple pulmonary nodules. Clin Chest Med. 1993;14:111–119. 20. Jennings SG, Winer-Muram HT, Tarver RD, et al. Lung tumor growth: assessment with CT—comparison of diameter and cross-sectional area with volume measurements. Radiology. 2004;231:866–871.
238 | www.thoracicimaging.com
Volume 27, Number 4, July 2012
21. Marten K, Auer F, Schmidt S, et al. Inadequacy of manual measurements compared to automated CT volumetry in assessment of treatment response of pulmonary metastases using RECIST criteria. Eur Radiol. 2006;16:781–790. 22. Yankelevitz DF, Reeves AP, Kostis WJ, et al. Small pulmonary nodules: volumetrically determined growth rates based on CT evaluation. Radiology. 2000;217:251–256. 23. Bolte H, Riedel C, Muller-Hulsbeck S, et al. Precision of computer-aided volumetry of artificial small solid pulmonary nodules in ex vivo porcine lungs. Br J Radiol. 2007;80:414–421. 24. Wormanns D, Kohl G, Klotz E, et al. Volumetric measurements of pulmonary nodules at multi-row detector CT: in vivo reproducibility. Eur Radiol. 2004;14:86–92. 25. Gietema HA, Wang Y, Xu D, et al. Pulmonary nodules detected at lung cancer screening: interobserver variability of semiautomated volume measurements. Radiology. 2006; 241:251–257. 26. Goodman LR, Gulsun M, Washington L, et al. Inherent variability of CT lung nodule measurements in vivo using semiautomated volumetric measurements. Am J Roentgenol. 2006;186:989–994. 27. Ashraf H, de HB, Shaker SB, et al. Lung nodule volumetry: segmentation algorithms within the same software package cannot be used interchangeably. Eur Radiol. 2010;20: 1878–1885. 28. Christe A, Torrente JC, Lin M, et al. CT screening and followup of lung nodules: effects of tube current-time setting and nodule size and density on detectability and of tube currenttime setting on apparent size. Am J Roentgenol. 2011;197: 623–630. 29. Gietema HA, Schaefer-Prokop CM, Mali WP, et al. Pulmonary nodules: Interscan variability of semiautomated volume measurements with multisection CT—influence of inspiration level, nodule size, and segmentation performance. Radiology. 2007;245:888–894. 30. Goo JM, Tongdee T, Tongdee R, et al. Volumetric measurement of synthetic lung nodules with multi-detector row CT: effect of various image reconstruction parameters and segmentation thresholds on measurement accuracy. Radiology. 2005;235:850–856. 31. Ko JP, Rusinek H, Jacobs EL, et al. Small pulmonary nodules: volume measurement at chest CT—phantom study. Radiology. 2003;228:864–870. 32. Ravenel JG, Leue WM, Nietert PJ, et al. Pulmonary nodule volume: effects of reconstruction parameters on automated measurements–a phantom study. Radiology. 2008;247:400–408. 33. Wang Y, van Klaveren RJ, van der Zaag-Loonen HJ, et al. Effect of nodule characteristics on variability of semiautomated volume measurements in pulmonary nodules detected in a lung cancer screening program. Radiology. 2008;248: 625–631. 34. Wang Y, de Bock GH, van Klaveren RJ, et al. Volumetric measurement of pulmonary nodules at low-dose chest CT: effect of reconstruction setting on measurement variability. Eur Radiol. 2010;20:1180–1187. 35. Xu DM, Gietema H, de KH, et al. Nodule management protocol of the NELSON randomised lung cancer screening trial. Lung Cancer. 2006;54:177–184. 36. Baldwin DR, Duffy SW, Wald NJ, et al. UK Lung Screen (UKLS) nodule management protocol: modelling of a single screen randomised controlled trial of low-dose CT screening for lung cancer. Thorax. 2011;66:308–313. 37. Hasegawa M, Sone S, Takashima S, et al. Growth rate of small lung cancers detected on mass CT screening. Br J Radiol. 2000;73:1252–1259. 38. Lindell RM, Hartman TE, Swensen SJ, et al. 5-year lung cancer screening experience: growth curves of 18 lung cancers compared to histologic type, CT attenuation, stage, survival, and size. Chest. 2009;136:1586–1595. 39. Nathan MH, Collins VP, Adams RA. Differentiation of benign and malignant pulmonary nodules by growth rate. Radiology. 1962;79:221–232. r
2012 Lippincott Williams & Wilkins
J Thorac Imaging
Volume 27, Number 4, July 2012
40. Revel MP, Merlin A, Peyrard S, et al. Software volumetric evaluation of doubling times for differentiating benign versus malignant pulmonary nodules. Am J Roentgenol. 2006;187: 135–142. 41. Winer-Muram HT, Jennings SG, Tarver RD, et al. Volumetric growth rate of stage I lung cancer prior to treatment: serial CT scanning. Radiology. 2002;223:798–805. 42. Yankelevitz DF, Kostis WJ, Henschke CI, et al. Overdiagnosis in chest radiographic screening for lung carcinoma: frequency. Cancer. 2003;97:1271–1275. 43. Lindell RM, Hartman TE, Swensen SJ, et al. Five-year lung cancer screening experience: CT appearance, growth rate, location, and histologic features of 61 lung cancers. Radiology. 2007;242:555–562. 44. Ko JP, Berman EJ, Kaur M, et al. Pulmonary nodules: growth rate assessment in patients by using serial CT and threedimensional volumetry. Radiology. 2012;262:662–671. 45. Ashraf H, Dirksen A, Loft A, et al. Combined use of positron emission tomography and volume doubling time in lung cancer screening with low-dose CT scanning. Thorax. 2011;66: 315–319. 46. Yankelevitz DF, Henschke CI. Does 2-year stability imply that pulmonary nodules are benign? Am J Roentgenol. 1997;168: 325–328. 47. Esmaili A, Munden RF, Mohammed TL. Small pulmonary nodule management: a survey of the members of the Society of Thoracic Radiology with comparison to the Fleischner Society guidelines. J Thorac Imaging. 2011;26:27–31. 48. Brandman S, Ko JP. Pulmonary nodule detection, characterization, and management with multidetector computed tomography. J Thorac Imaging. 2011;26:90–105. 49. Xu DM, van Klaveren RJ, de Bock GH, et al. Limited value of shape, margin and CT density in the discrimination between benign and malignant screen detected solid pulmonary nodules of the NELSON trial. Eur J Radiol. 2008;68:347–352. 50. Xu DM, van der Zaag-Loonen HJ, Oudkerk M, et al. Smooth or attached solid indeterminate nodules detected at baseline CT screening in the NELSON study: cancer risk during 1 year of follow-up. Radiology. 2009;250:264–272. 51. Ahn MI, Gleeson TG, Chan IH, et al. Perifissural nodules seen at CT screening for lung cancer. Radiology. 2010;254: 949–956. 52. Bankoff MS, McEniff NJ, Bhadelia RA, et al. Prevalence of pathologically proven intrapulmonary lymph nodes and their appearance on CT. Am J Roentgenol. 1996;167: 629–630. 53. Hyodo T, Kanazawa S, Dendo S, et al. Intrapulmonary lymph nodes: thin-section CT findings, pathological findings, and CT
r
2012 Lippincott Williams & Wilkins
Management of Pulmonary Nodules
54.
55. 56.
57.
58. 59.
60.
61.
62. 63.
64.
65.
differential diagnosis from pulmonary metastatic nodules. Acta Med Okayama. 2004;58:235–240. Ishikawa H, Koizumi N, Morita T, et al. Ultrasmall intrapulmonary lymph node: usual high-resolution computed tomographic findings with histopathologic correlation. J Comput Assist Tomogr. 2007;31:409–413. Shaham D, Vazquez M, Bogot NR, et al. CT features of intrapulmonary lymph nodes confirmed by cytology. Clin Imaging. 2010;34:185–190. Silvestri GA, Gould MK, Margolis ML, et al. Noninvasive staging of non-small cell lung cancer: ACCP evidenced-based clinical practice guidelines (2nd edition). Chest. 2007;132 (suppl 3):178S–201S. Gould MK, Maclean CC, Kuschner WG, et al. Accuracy of positron emission tomography for diagnosis of pulmonary nodules and mass lesions: a meta-analysis. JAMA. 2001; 285:914–924. Gould MK, Sanders GD, Barnett PG, et al. Cost-effectiveness of alternative management strategies for patients with solitary pulmonary nodules. Ann Intern Med. 2003;138:724–735. United Nations Scientific Committee on the Effects of Atomic Radiation. Sources and Effects of Ionizing Radiation: United Nations Scientific Committee on the Effects of Atomic Radiation UNSCEAR 2000 Report to the General Assembly, With Scientific Annexes. New York, NY: United Nations; 2000. Yi CA, Lee KS, Kim BT, et al. Tissue characterization of solitary pulmonary nodule: comparative study between helical dynamic CT and integrated PET/CT. J Nucl Med. 2006; 47:443–450. Veronesi G, Bellomi M, Mulshine JL, et al. Lung cancer screening with low-dose computed tomography: a non-invasive diagnostic protocol for baseline lung nodules. Lung Cancer. 2008;61:340–349. Swensen SJ, Viggiano RW, Midthun DE, et al. Lung nodule enhancement at CT: multicenter study. Radiology. 2000;214: 73–80. 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;233:191–199. Wang Y, van Klaveren RJ, de Bock GH, et al. No benefit for consensus double reading at baseline screening for lung cancer with the use of semiautomated volumetry software. Radiology. 2012;262:320–326. Infante M, Cavuto S, Lutman FR, et al. A randomized study of lung cancer screening with spiral computed tomography: three-year results from the DANTE trial. Am J Respir Crit Care Med. 2009;180:445–453.
www.thoracicimaging.com |
239