Chest Imaging
Original Research
Management of Patients with Small Pulmonary Nodules: A Survey of Radiologists, Pulmonologists, and Thoracic Surgeons
Audio Available | 
Share
OBJECTIVE. The objective of our study was to survey recommendations for the management of small pulmonary nodules found on helical CT from radiologists, pulmonologists, and thoracic surgeons.
MATERIALS AND METHODS. We surveyed 774 radiologists, 623 pulmonologists, and 101 thoracic surgeons. All are members of an associated Austrian scientific society and were asked for their recommendations in 13 hypothetical cases. Statistical analysis was focused on possible differences between medical specialities, residents and fellows, and male and female doctors and on a possible influence of the number of years in training or in medical practice.
RESULTS. Complete surveys were returned from 91 radiologists, 74 pulmonologists, and 12 thoracic surgeons. The most frequent recommendation from radiologists was follow-up, whereas the most frequent recommendation from pulmonologists and surgeons was biopsy. The most frequently advised CT follow-up period was 3 months. Whereas thoracic surgeons preferred video-assisted thoracoscopy, radiologists and pulmonologists recommended less invasive procedures (bronchoscopy, CT-guided biopsy) to gain a specific diagnosis.
CONCLUSION. There are significant differences in the management of small pulmonary nodules among radiologists, pulmonologists, and thoracic surgeons. Whereas radiologists tend to recommend short-term follow-up, pulmonologists and thoracic surgeons prefer a more aggressive approach, especially in patients with a higher likelihood of malignancy.
Keywords: chest, lung cancer, MDCT, pneumatology, pulmonology, pulmonary nodules
| Introduction |   |
|---|
With the improved quality and increasing use of MDCT, small pulmonary nodules— that is, nodules smaller than 1 cm in diameter—are frequently detected on chest CT scans [1, 2]. The debate over the appropriate management of small pulmonary nodules is ongoing. In individual patients, the diagnostic or therapeutic consequence of one or more small pulmonary nodules detected on MDCT depends on the content of the individual radiologic report and the interpretation of the report by the referring physician. In 2001, Munden and Hess [3] published a survey administered to the members of the Society of Thoracic Radiology regarding how to manage patients with small pulmonary nodules. That study offered valuable insights into the range of opinions regarding this topic within that radiologic society. The study showed that most radiologists preferred to follow the patient with short-term follow-up rather than to advise biopsy. However, the study did not investigate how referring physicians rate small pulmonary nodules when noted in a CT report.
In this study, we surveyed the members of the Austrian Radiological Society, the Austrian Society of Pneumonology, and the Austrian Society for Thoracic and Cardiovascular Surgery regarding their opinions on the management of small pulmonary nodules.
| Materials and Methods | |
|---|
In March 2004, we mailed a questionnaire regarding small pulmonary nodules on chest CT to 774 listed members of the Austrian Radiological Society, 623 members of the Austrian Society of Pneumonology, and 101 members of the Austrian Society for Thoracic and Cardiovascular Surgery. The members of the societies were asked for their management decisions in 13 hypothetical cases with small pulmonary nodules. The questionnaire was distributed with a cover letter stating the purpose of the study and the support of the three scientific societies. In the introduction section of the questionnaire, small pulmonary nodules were defined as solid intrapulmonary lesions without calcifications that were 3-5 mm, had sharp or ill-defined margins, and were in any location in the lung.
To allow comparison of our study population with American radiologists, the survey questions followed the case scenarios published by Munden and Hess [3] (Appendix 1). The participants were asked to choose one from nine possible answers, as shown in Appendix 1, or to give one alternative answer. In contrast to Munden and Hess, we asked participants to choose among three follow-up intervals (3, 6, or 12 months) and three biopsy techniques (bronchoscopy, CT-guided biopsy, or video-assisted thoracoscopy). These modifications were made to allow accurate answers regarding diagnostic tools that are exclusively in clinical use within a specific group of specialists in our country.
In addition to the case scenarios, the participating colleagues were asked demographic information regarding age, sex, medical speciality, years of training or clinical experience, working place (hospital, private practice, or both) and whether they worked in a referral center for lung cancer. Because tuberculosis is homogeneously distributed in Austria and histoplasmosis is not endemic, specific questions regarding the geographic location of the responders were not included [4].
Statistical analysis was performed using SAS software (version 9, SAS Institute). To determine the association of the answers given and the responder's specialty, experience, sex, and whether the responder worked in a lung cancer reference center, exact p values were calculated. A p value below 0.05 was considered significant. Spearman's rank correlation was used to describe the association between subsequent answers.
| Results | |
|---|
One hundred eighty questionnaires were returned. Three questionnaires had to be excluded because they were incomplete. Of the remaining 177 questionnaires, 91 were returned from radiologists, 74 from pulmonologists, and 12 from thoracic surgeons. The population characteristics of the three groups are given in Table 1.
Table 2 summarizes the answers for each case separated by the responders' specialty. Radiologists—The most frequent recommendation from radiologists was follow-up, advised in all but two cases. In case 4 (patient with lung cancer and a small pulmonary nodule in the contralateral lung), an almost equal number of radiologists favored follow-up as favored biopsy (43% vs 44%). In case 10 (patient with a known extrathoracic malignancy with more than three small pulmonary nodules), the most frequent recommendation from radiologists was biopsy.
Pulmonologists—The most frequent recommendation from pulmonologists was biopsy, advised in eight of the 13 cases. In the remaining five cases, follow-up was the most frequent recommendation. In case 4 (patient with lung cancer and a small pulmonary nodule in the contralateral lung), 39% of the pulmonologists did not choose one of the offered answers but mainly preferred to further evaluate the patient with PET.
Thoracic surgeons—Thoracic surgeons preferred biopsy in nine cases and follow-up in four cases.
The differences in the recommendations among the three medical subspecialties reached statistical significance in all cases except case 5 (patient with a resected [R0] non-small cell lung cancer in whom one small pulmonary nodule is found on follow-up CT).
In all but two cases (cases 1 and 2), radiologists preferred a 3-month follow-up interval. Most of the pulmonologists preferred a 3-month follow-up interval in all cases.
Surgeons recommended, in cases 7 and 10, a 6-month follow-up interval; in all other cases, they preferred a 3-month follow-up.
If radiologists advised biopsy, the most frequently advised techniques were bronchoscopically guided biopsy in six cases, CT-guided biopsy in three cases, and video-assisted thoracoscopy (VATS) in four cases. Pulmonologists preferred bronchoscopically guided biopsy in 11 cases and CT-guided biopsy in one case. In one case (case 1), no single pulmonologist recommended biopsy. In contrast, thoracic surgeons recommended biopsy by VATS in all cases.
Within the three groups of specialists, no significant differences between the answers of male and female physicians were observed.
To assess the influence of personal clinical experience on their recommendations, we grouped responders as follows: residents in training, physicians with up to 10 years of clinical experience, and physicians with more than 10 years of clinical experience.
Among the radiologists, experience proved to be of only minor significance on the recommendations. Differences in the recommendations were statistically significant only for cases 4 and 10. In case 4, a 3-month follow-up was recommended by 41% of the radiologists in training and by 34% of the radiologists with more than 10 years of experience. In case 10, 32% of radiologists in training recommended a 3-month follow-up interval in contrast to radiologists with more than 10 years of experience, whose most frequent answer was biopsy using VATS.
Among the pulmonologists, experience had a significant influence on the recommendations in cases 4, 6, 8, and 13. For example, in case 6 (one small pulmonary nodule in a patient with a new diagnosis of extrathoracic malignancy), pulmonologists in training recommended CT-guided biopsy (50%), whereas a 3-month follow-up interval was the most frequent answer of pulmonologists with less and more than 10 years of clinical practice (26% and 44%, respectively).
Among the thoracic surgeons, differences between the three groups of experience did not reach statistical significance.
Working in a lung cancer referral center did not have a significant influence on the recommendations of responders of all three medical specialities. In addition, differences between radiologists working in private practice and those working in a hospital did not reach statistical significance.
Spearman's rank correlation did not show any correlation between the recommendations given—that is, we observed that physicians evaluated each case individually and did not tend to express a consistent recommendation fitting all individual cases.
| Discussion | |
|---|
The results of this survey clearly show significant differences among the radiology, pulmonology, and thoracic surgery communities in Austria, a middle European country with a population of about 8 million people and a high-end public health care system. In the next years, diagnostic and therapeutic problems with small pulmonary nodules might quantitatively increase because spatial resolution of lung CT is further increased by the routine use of MDCT scanners. Screening projects using CT for the early detection of lung cancer found pulmonary nodules in 23-74% of a population at risk [5-7], depending on the CT technology used. Most of these nodules, however, were benign [7]. Even in children with solid malignancies, a significant number of pulmonary nodules are present, most of which are benign [8]. During the past several years, CT criteria for the differentiation of benign versus malignant pulmonary nodules have been adopted. Certain morphologic criteria, such as a solid or part-solid density, are more suspicious for malignancy than others [9]. Nevertheless, there is general agreement that size correlates with malignancy, but CT morphology does not reliably allow differentiation between benign and malignant small pulmonary nodules [9]. In this setting, besides personal experience and individual preference, the clinical pretest probability for malignancy, as addressed in our different case scenarios, and guidelines about patient management might influence an individual physician.
Over the past years, guidelines for the management of pulmonary nodules smaller than 3 cm but larger than 1 cm have been developed [10, 11]. However, the management of very small nodules (3-5 mm in diameter) remains a diagnostic dilemma. Small pulmonary nodules are usually detected by radiologists who report their findings to the referring physicians. In addition to the findings, the radiologic report should contain recommendations for further diagnostic workup [12]. Decisions regarding further management are therefore frequently influenced by the recommendations of radiologists. Unfortunately, recommendations are only rarely included in a radiologic report, as shown in a study evaluating CT and CT reports in patients with lung cancer in a tertiary referral center (Gahleitner-Schick S et al., presented at the 2004 annual meeting of the Radiologic Society of North America). Consequently, the actual patient management depends exclusively on the referring physician. In an ideal world, there should be consensus of different medical specialists about how to manage a patient with one or more small pulmonary nodules.
A survey among the members of the Society of Thoracic Radiology by Munden and Hess [3] showed that radiologists in the United States tend to recommend short-term follow-up (3-6 months) if they detect small pulmonary nodules on CT scans [3]. Interestingly, the Austrian radiologists surveyed in this study also recommended short-term follow-up more often than other approaches in these scenarios. The most frequently advised interval for follow-up was 3 months for all three specialities. This approach is consistent with the 1999 guidelines of the Early Lung Cancer Action Project for the evaluation of nodules with a diameter of 5 mm or less [5]. For nodules with diameters between 6 and 10 mm, the same protocol recommended CT-guided biopsy or VATS for further evaluation as a possible management strategy [5].
In a recent evaluation [11] of the data from that study [5], published shortly after the distribution of our questionnaires, the authors now propose a CT follow-up interval of 1 year in patients with small pulmonary nodules < 5 mm or nonsolid nodules 5-9 mm in diameter. For part-solid or solid nodules with a diameter between 5 and 9 mm, they suggest a follow-up interval of 6-12 weeks to assess for possible rapid nodule growth or resolution [11]. The recommendation for the 1-year follow-up interval for solid nodules < 5 mm in this protocol results from the fact that lung cancer is very rare in nodules of this size [11]. The assessment of small pulmonary nodule growth (having average tumor doubling times of non-small cell lung carcinomas in mind) is of limited reliability if one measures nodule diameters rather than using volumetric software [13]. However, to the best of our knowledge, volumetric software is not in clinical use in Austria except for research purposes.
Our study results show there are significant differences in the management of patients with small pulmonary nodules among radiologists, pulmonologists, and thoracic surgeons in Austria. In contrast to the radiology community, pulmonologists and thoracic surgeons more frequently recommend determining a specific tissue diagnosis, especially in patients with a higher likelihood of malignancy. Why these specialities more aggressively approach small pulmonary nodules remains speculative. Direct urging from patients for a definitive diagnosis instead of uncertainty throughout the follow-up, easy access to invasive techniques, and reimbursement for these procedures are possible reasons.
These recommendations have to be seen in light of the accuracy of the specific technique for tissue diagnosis or for differentiation between benign and malignant nodules. The accuracy of bronchoscopy in the evaluation of small pulmonary nodules is limited [14, 15]. A drop in accuracy with decreasing lesion size has been noted for CT-guided biopsy as well, even if modern techniques such as fluoroscopy are used [16]. For nodules that are 3-5 mm, specific data regarding accuracy are not available, but an even lower accuracy for CT-guided biopsy of lesions less than 2 cm has to be expected [16]. VATS, on the other hand, is a reliable method to obtain diagnosis and get rid of the small pulmonary nodule. However, compared with CT-guided biopsy and bronchoscopy, VATS is the most invasive and expensive technique with low, but significant, morbidity [17]. Thus, if a tissue diagnosis in a patient with a small pulmonary nodule is considered, VATS should be performed only in patients with a high pretest probability for malignancy. Regarding 18F-FDG PET, sensitivity and specificity are limited in the evaluation of pulmonary nodules that are about 5 mm [18]. Regarding our definition of small pulmonary nodules (3-5 mm), one can assume that the very small size may not have been accounted for by all responders who suggested invasive management.
Differences between male and female physicians did not reach statistical significance. However, differences between residents and experienced physicians in the pulmonology, but not in the radiology, community reached statistical significance. One explanation might be that the radiology community is more uniformly involved in CT than pulmonologists are informed about imaging issues.
There are definitely some drawbacks to this study. First, the number of questionnaires returned was only about 10% in all groups of specialists. This might in part be explained by the fact that the Austrian Radiological Society is the scientific home of all Austrian radiologists, not only those dedicated to chest radiology. Obviously, small pulmonary nodules on CT scans did not reach high interest within the members of the other two societies. Because of the low number of thoracic surgeons in our country, the low response rate might have influenced possible significant differences in this group. Nevertheless, surgeons presented with the highest homogeneity in their approaches to the clinical scenarios, making their results meaningful.
Second, we did not survey other medical specialists who might be involved in the management of these patients. Precisely, this might have been oncologists and general practitioners, who are definitely involved in some of the presented clinical scenarios.
Third, one might argue that the results obtained in a small country with a public health care system might differ from other communities with different health care systems. This might be true, but in our opinion the approach to small pulmonary nodules detected on CT might be comparable to all other communities where access to and distribution of high-end MDCT are comparable to Austria.
In conclusion, the presented data confirm large variability in the management of patients with small pulmonary nodules among radiologists, pulmonologists, and thoracic surgeons. Further, diagnostic strategies within the three groups of specialists seem to be based on individual access to specific techniques and on clinical experiences with the typical pulmonary nodule, which measures 1-3 cm. Given the high number of benign small pulmonary nodules detected on CT-based screening studies and the low accuracy of CT-guided biopsy and even lower accuracy of bronchoscopy in the tissue diagnosis of lesions about 5 mm in size, these data show the need for an interdisciplinary consensus and interdisciplinary guidelines for the management of these patients.
| APPENDIX 1: Possible Answers, Background Questions, and Case Scenarios | |
|---|
Answer 1: I don't do anything. The finding will not be reported.
Answer 2: I don't do anything. This is certainly a benign process. The finding will be reported; however, no further procedure is suggested.
Answer 3: CT, with follow-up in 3 months.
Answer 4: CT, with follow-up in 6 months.
Answer 5: CT, with follow-up in 12 months.
Answer 6: Bronchoscopy.
Answer 7: Biopsy (CT-guided).
Answer 8: Video-assisted thoracoscopy (VATS).
Answer 9: Certainly malign; no biopsy, eventual surgery.
Answer 10: Something different,....
Speciality
Years of training/practice
Do you work in a hospital, private practice, or both?
Is your main practice at a referral center for lung diseases?
In a young patient (< 40 years) with no previous malignancy, one small pulmonary nodule is found on chest CT.
aCase scenarios are taken from [3].
In a patient older than 40 years with no previous malignancy, one small pulmonary nodule is found on chest CT.
In patient considered at high risk for lung cancer with no previous malignancy, one small pulmonary nodule is found on chest CT.
In a patient with a verified resectable lung cancer, one small pulmonary nodule is found in the contralateral lung.
In a patient with a resected (R0) non-small cell lung cancer, one small pulmonary nodule is found in a follow-up CT.
In a patient with a new diagnosis of extrathoracic malignancy, one small pulmonary nodule is found in the staging CT.
In a patient with a history of extrathoracic malignancy, one small pulmonary nodule is found on chest CT.
In a young patient (< 40 years) with no previous malignancy, some (n > 3) small pulmonary nodules are found on chest CT.
In a patient older than 40 years with no previous malignancy, some (n > 3) small pulmonary nodules are found on chest CT.
In a patient with a known extrathoracic malignancy, some (n > 3) small pulmonary nodules are found on chest CT.
In a patient with no previous malignancy, CT reveals some calcified nodules and one small pulmonary nodule.
In a patient with a known malignancy, CT reveals some calcified nodules and one small pulmonary nodule.
In a patient with a high risk for lung cancer, CT reveals some calcified nodules and one small pulmonary nodule.
Supported by the Austrian Radiological Society, the Austrian Society of Pneumonology, the Austrian Society for Thoracic and Cardiovascular Surgery, and the Akademie für bildgebende Diagnostik und radiologisch-interventionelle Therapie (ADRAT).
Address correspondence to G. H. Mostbeck ([email protected]).
We thank the Austrian Radiological Society, the Austrian Society of Pneumonology, the Austrian Society for Thoracic and Cardiovascular Surgery, and the Akademie für bildgebende Diagnostik und radiologisch-interventionelle Therapie (ADRAT).
| References | |
|---|
