Evaluation and Management of Small Pulmonary Nodules
Evaluation and Management of Small Pulmonary Nodules
Initial evaluation consists of a detailed history, physical exam and review of old and new images, as well as a chest CT scan with thin 1 mm slices through the lung nodules.
Multiple models are available to estimate the pretest probability of malignant pulmonary nodules. These models were validated for evaluation of solitary pulmonary nodules. In this article, we mention the Mayo clinic model and the Veterans Affairs model.
The Mayo Clinic model is defined by the equations:
where SPN is the solitary pulmonary nodule, e is the base of the natural logarithm, age indicates the patient's age in years; smoke indicates smoking history (1 = current or former smoker, 0 = never smoker); cancer indicates history of an extrathoracic cancer 5 or more years before nodule identification (1 = yes, 0 = no or not specified); diameter indicates the largest nodule measurement, in mm, reported on initial chest x-ray or CT; spiculation indicates mention of nodule spiculation on any imaging test report (1 = yes, 0 = no or not specified); and upper is location of the nodule within the upper lobe of either lung (1 = yes, 0 = no).
The Veterans Affairs model is defined by the equations:
where e is the base of the natural logarithm; smoke indicates smoking history (1 = current or former smoker, and 0 = never smoker); age 10 indicates age in years at the time of nodule identification, divided by 10; diameter indicates the largest nodule measurement, in mm, reported on initial chest x-ray or CT; and years-quit 10 indicates the number of years since quitting smoking, divided by 10 (0 indicates not applicable).
Lesions that are highly suspicious for malignancy based on these prediction models, lesions >30 mm in size or a documented increase in size are very suspicious for malignancy and should be considered for tissue diagnosis. Benign nodules that have a benign pattern of calcification or have remained stable in volume over the 2-year period in a young nonsmoker patient do not need further work up. However, after completion of the initial evaluation most lung nodules will probably be termed indeterminate and will require further diagnostic evaluation. These evaluations will range from careful observation to PET scanning to biopsy.
Observation is the least invasive approach in the evaluation of lung nodules preventing the patients from exposure to the risk of invasive procedures and surgery. Such observation may be offered to patients with very low risk of malignancy (based on clinical and radiological risk factors). In addition, patients with an exceedingly high risk of complications from invasive procedure due to comorbid pulmonary or cardiac condition and/or bleeding tendency may be offered such an approach.
Many of the lung cancer screening trials have used a rather aggressive approach to radiographic follow-up of indeterminate nodules and consisted of re-imaging at 3, 6, 12, 18 and 24 months after the initial scan. An approach advocated by the American College of Chest Physicians is slightly less aggressive advising not to have additional scanning between 12 and 24 months (i.e., at 3, 6, 12 and 24 months) unless the patient has a higher than average risk of malignancy. The Fleischner Society for Thoracic Imaging and Diagnosis recommendations went a bit further by risk-stratifying patients and nodules based on clinical and radiological risk factors previously mentioned and specific size range, and combining both to advise specific frequency of radiological follow-up ( Table 2 ). Detection of no change in nodule size over a 2-year period continues to be an indication to stopping the follow-up unless the nodules have ground-glass attenuation, in which case a longer follow-up may be considered. A new statement published by the Fleischner Society regarding the appropriate follow-up of pulmonary nodules distinguished between pure ground-glass nodules (GGN) and partly GGN that have a solid component. The new recommendations emphasize the importance of obtaining an initial follow-up at 3 months, given that many GGN may resolve within 3 months. Persistent nodules need to be followed up depending on the size and on whether it is solitary nodule or multiple nodules. For purely GGN the recommended follow-up is for a minimum of 3 years based on the fact that GGN carry a higher risk for premalignant lesions or adenocarcinoma. This risk is greater in part-solid GGN where the recommendation is to biopsy or resection especially when the solid component is greater than 5 mm. PET scans are not recommended in the evaluation of GGN given its limited value and misleading results. PET scans may be considered for solitary part-solid GGN >10 mm in size. The recommendations are summarized in Table 3 .
Considering a second approach to indeterminate lung nodules, a patient may be offered a further characterization of the nodules with PET scans.
Performing a PET scan has been shown to be cost effective in: patients with an apparent low risk of harboring a malignancy and suspicious radiological features of a lung nodule; and patients with an exceedingly high risk of surgical morbidity and mortality or those reluctant to undergo invasive testing. If PET results are negative in these cases, further follow-up with repeated CT scans would be recommended. Another benefit of PET scanning is its ability to detect occult metastasis to mediastinal lymph nodes and/or other organs that may direct biopsy to metastasis sites to establish staging. False-negative PET scan results increase with nodules <10 mm in size.
The third approach to management of indeterminate lung nodules is to obtain tissue diagnosis by fiberoptic bronchoscopy (FOB), transthoracic needle aspiration (TTNA) or surgical resection with video-assisted thoracoscopy (VATS) and thoracotomy. FOB has an overall diagnostic yield that ranges between 36 and 68%. Diagnostic yield is somewhat higher for malignant lesions ranging from 44 to 68% and lower for benign lesions (12–41%). Several factors that may increase diagnostic yield are presence of symptoms (cough, hemoptysis and local wheeze), nodule size >2 cm (yield 56%), location of nodule within the inner two-thirds of a lung field and presence of a bronchus sign, where a segment of a bronchus is seen on CT passing through or leading directly towards the nodule of interest. When this particular sign is present the diagnostic yield improves to 60–90% compared with 14–30% when it is absent. Diagnostic interventions that could be useful in the diagnosis of lung nodules by FOB include bronchial washings, brushing, endobronchial biopsy, transbronchial needle aspiration and endobronchial ultrasonography with needle aspiration. More recently a novel approach involving electromagnetic guidance to the nodule has emerged. This technology utilizes an electromagnetic board and an endobronchial position sensor to create real-time reconstruction and overlay of previously acquired multiplanar CT images. The resulting computer model of a bronchial tree is then used to guide, in real life, an extended working channel in close proximity to the nodule of interest where an endobronchial biopsy or transbronchial needle aspiration may be performed. Reported diagnostic yield of this technique varies from 67 to 85% and appears to be highly dependent on the aforementioned bronchus sign. The risk of pneumothorax appears to be comparable to normal transbronchial biopsy.
Another evolving modality for guidance is endobronchial ultrasound. With this modality a thin ultrasound probe is introduced through the scope's working channel and pushed with the scope towards the supposed location of the abnormality. Once the lesion is identified on the ultrasound monitor the probe is withdrawn and biopsy forceps used to obtain the specimen. A modification to this technique is the use of a guidance channel covering the ultrasound probe that stays in the working channel while the probe is removed. This provides a direct pathway for the forceps to the site of biopsy. In one study such an approach resulted in an overall diagnostic yield of 77%; however, this was highly dependent on location of the probe, within versus adjacent to the lesion (76 vs 46%, respectively). The size of the lesion did not appear to significantly affect the diagnostic yield, varying from 77% in lesions 3 cm or less to 76% in lesions 1 cm or less. A recently performed meta-analysis evaluated the utility of endobronchial radial ultrasonography (EBUS). It showed that EBUS had point specificity of 1.00 (95% CI: 0.99–1.00) and point sensitivity of 0.73 (95% CI: 0.70–0.76) for the detection of lung cancer, with a positive likelihood ratio of 26.84 (12.60–57.20) and a negative likelihood ratio of 0.28 (0.23–0.36). Significant interstudy heterogeneity for sensitivity was observed and was likely dependent on the prevalence of malignancy in the study population and lesion size.
An overall utility of any guiding method during FOB in a work up of a pulmonary nodule was assessed in another recent meta-analysis. Authors have concluded that a diagnostic yield of guided bronchoscopic techniques is better than the traditional transbronchial approach. Although the yield remains lower than TTNA, the procedural risk appears to be lower.
Most recently another method for guided biopsy has emerged. Virtual bronchosocopy (VB) is a method of guidance where CT scanning is used to make a map of the bronchial tree, which the physician can then view on a screen while navigating the real bronchial tree with an ultra thin bronchoscope. VB differs from electromagnetic navigation in that it does not rely on real-time positional signaling from sensors around the patient. Ishida et al. used several medical centers in Japan to randomize 199 patients with peripheral lesions <3 cm to undergo a biopsy using either VB-assisted endobronchial ultrasound (VB + EBUS) or EBUS alone. A pathological diagnosis was achieved in 80% of patients in the VB + EBUS group, but only 67% of those in the EBUS alone group (p = 0.032). This is the first study that directly and prospectively evaluated the value of VB as an adjunct to EBUS-guided biopsy. Given that the evidence for VB and VB + EBUS is not nearly as robust as the evidence available for EBUS alone and/or combination of EBUS and electromagnetic navigation, it would be prudent to limit the use of this technology until more data is available.
TTNA or core biopsy, best performed under CT guidance is another useful procedure to diagnose indeterminate solitary pulmonary nodules, particularly with peripheral lesions, and can be performed on nodules less than 2 cm in size. The diagnostic accuracy of this procedure exceeds 95% for malignant lesions with slightly poorer performance for benign lesions. The diagnostic yield of TTNA can be improved even further when core biopsy is done and the cytopathologist is available to determine the adequacy of the specimen. This procedure does carry risks, including hemorrhage and pneumothorax. Pneumothorax is a much more frequent complication occurring in 10–35% of cases with only 5–10% of those patients requiring chest tube placement. In a more recent, population-based cross-sectional analysis of over 15,000 patients who underwent TTNA, hemorrhage was reported in 1% and pneumothorax in 15% with 6.6% requiring chest tube placement. Significant risk factors for those patients with complications appeared to be between 60 and 69 years of age, current smokers and those with a diagnosis of chronic obstructive pulmonary disease. Therefore, TTNA should be considered with caution in patients with poor lung reserve and postpneumonectomy.
Use of ultrasound for guidance of a needle, especially in peripheral nodules, offers a radiation-free alternative to CT-guided biopsy. The diagnostic accuracy appears to be comparable with a CT-guided approach with 92% for malignant lesions, a negative predictive value of 93% and an overall complication rate of 10% in some studies. While this approach is feasible and has been proven effective even for lesions less than 1 cm in size, great care must be taken to select appropriate patients and lesions, as well as assurance of appropriate experience of the operator.
Approach to the Management of Pulmonary Nodules
Initial evaluation consists of a detailed history, physical exam and review of old and new images, as well as a chest CT scan with thin 1 mm slices through the lung nodules.
Multiple models are available to estimate the pretest probability of malignant pulmonary nodules. These models were validated for evaluation of solitary pulmonary nodules. In this article, we mention the Mayo clinic model and the Veterans Affairs model.
The Mayo Clinic model is defined by the equations:
where SPN is the solitary pulmonary nodule, e is the base of the natural logarithm, age indicates the patient's age in years; smoke indicates smoking history (1 = current or former smoker, 0 = never smoker); cancer indicates history of an extrathoracic cancer 5 or more years before nodule identification (1 = yes, 0 = no or not specified); diameter indicates the largest nodule measurement, in mm, reported on initial chest x-ray or CT; spiculation indicates mention of nodule spiculation on any imaging test report (1 = yes, 0 = no or not specified); and upper is location of the nodule within the upper lobe of either lung (1 = yes, 0 = no).
The Veterans Affairs model is defined by the equations:
where e is the base of the natural logarithm; smoke indicates smoking history (1 = current or former smoker, and 0 = never smoker); age 10 indicates age in years at the time of nodule identification, divided by 10; diameter indicates the largest nodule measurement, in mm, reported on initial chest x-ray or CT; and years-quit 10 indicates the number of years since quitting smoking, divided by 10 (0 indicates not applicable).
Lesions that are highly suspicious for malignancy based on these prediction models, lesions >30 mm in size or a documented increase in size are very suspicious for malignancy and should be considered for tissue diagnosis. Benign nodules that have a benign pattern of calcification or have remained stable in volume over the 2-year period in a young nonsmoker patient do not need further work up. However, after completion of the initial evaluation most lung nodules will probably be termed indeterminate and will require further diagnostic evaluation. These evaluations will range from careful observation to PET scanning to biopsy.
Observation is the least invasive approach in the evaluation of lung nodules preventing the patients from exposure to the risk of invasive procedures and surgery. Such observation may be offered to patients with very low risk of malignancy (based on clinical and radiological risk factors). In addition, patients with an exceedingly high risk of complications from invasive procedure due to comorbid pulmonary or cardiac condition and/or bleeding tendency may be offered such an approach.
Many of the lung cancer screening trials have used a rather aggressive approach to radiographic follow-up of indeterminate nodules and consisted of re-imaging at 3, 6, 12, 18 and 24 months after the initial scan. An approach advocated by the American College of Chest Physicians is slightly less aggressive advising not to have additional scanning between 12 and 24 months (i.e., at 3, 6, 12 and 24 months) unless the patient has a higher than average risk of malignancy. The Fleischner Society for Thoracic Imaging and Diagnosis recommendations went a bit further by risk-stratifying patients and nodules based on clinical and radiological risk factors previously mentioned and specific size range, and combining both to advise specific frequency of radiological follow-up ( Table 2 ). Detection of no change in nodule size over a 2-year period continues to be an indication to stopping the follow-up unless the nodules have ground-glass attenuation, in which case a longer follow-up may be considered. A new statement published by the Fleischner Society regarding the appropriate follow-up of pulmonary nodules distinguished between pure ground-glass nodules (GGN) and partly GGN that have a solid component. The new recommendations emphasize the importance of obtaining an initial follow-up at 3 months, given that many GGN may resolve within 3 months. Persistent nodules need to be followed up depending on the size and on whether it is solitary nodule or multiple nodules. For purely GGN the recommended follow-up is for a minimum of 3 years based on the fact that GGN carry a higher risk for premalignant lesions or adenocarcinoma. This risk is greater in part-solid GGN where the recommendation is to biopsy or resection especially when the solid component is greater than 5 mm. PET scans are not recommended in the evaluation of GGN given its limited value and misleading results. PET scans may be considered for solitary part-solid GGN >10 mm in size. The recommendations are summarized in Table 3 .
Considering a second approach to indeterminate lung nodules, a patient may be offered a further characterization of the nodules with PET scans.
Performing a PET scan has been shown to be cost effective in: patients with an apparent low risk of harboring a malignancy and suspicious radiological features of a lung nodule; and patients with an exceedingly high risk of surgical morbidity and mortality or those reluctant to undergo invasive testing. If PET results are negative in these cases, further follow-up with repeated CT scans would be recommended. Another benefit of PET scanning is its ability to detect occult metastasis to mediastinal lymph nodes and/or other organs that may direct biopsy to metastasis sites to establish staging. False-negative PET scan results increase with nodules <10 mm in size.
The third approach to management of indeterminate lung nodules is to obtain tissue diagnosis by fiberoptic bronchoscopy (FOB), transthoracic needle aspiration (TTNA) or surgical resection with video-assisted thoracoscopy (VATS) and thoracotomy. FOB has an overall diagnostic yield that ranges between 36 and 68%. Diagnostic yield is somewhat higher for malignant lesions ranging from 44 to 68% and lower for benign lesions (12–41%). Several factors that may increase diagnostic yield are presence of symptoms (cough, hemoptysis and local wheeze), nodule size >2 cm (yield 56%), location of nodule within the inner two-thirds of a lung field and presence of a bronchus sign, where a segment of a bronchus is seen on CT passing through or leading directly towards the nodule of interest. When this particular sign is present the diagnostic yield improves to 60–90% compared with 14–30% when it is absent. Diagnostic interventions that could be useful in the diagnosis of lung nodules by FOB include bronchial washings, brushing, endobronchial biopsy, transbronchial needle aspiration and endobronchial ultrasonography with needle aspiration. More recently a novel approach involving electromagnetic guidance to the nodule has emerged. This technology utilizes an electromagnetic board and an endobronchial position sensor to create real-time reconstruction and overlay of previously acquired multiplanar CT images. The resulting computer model of a bronchial tree is then used to guide, in real life, an extended working channel in close proximity to the nodule of interest where an endobronchial biopsy or transbronchial needle aspiration may be performed. Reported diagnostic yield of this technique varies from 67 to 85% and appears to be highly dependent on the aforementioned bronchus sign. The risk of pneumothorax appears to be comparable to normal transbronchial biopsy.
Another evolving modality for guidance is endobronchial ultrasound. With this modality a thin ultrasound probe is introduced through the scope's working channel and pushed with the scope towards the supposed location of the abnormality. Once the lesion is identified on the ultrasound monitor the probe is withdrawn and biopsy forceps used to obtain the specimen. A modification to this technique is the use of a guidance channel covering the ultrasound probe that stays in the working channel while the probe is removed. This provides a direct pathway for the forceps to the site of biopsy. In one study such an approach resulted in an overall diagnostic yield of 77%; however, this was highly dependent on location of the probe, within versus adjacent to the lesion (76 vs 46%, respectively). The size of the lesion did not appear to significantly affect the diagnostic yield, varying from 77% in lesions 3 cm or less to 76% in lesions 1 cm or less. A recently performed meta-analysis evaluated the utility of endobronchial radial ultrasonography (EBUS). It showed that EBUS had point specificity of 1.00 (95% CI: 0.99–1.00) and point sensitivity of 0.73 (95% CI: 0.70–0.76) for the detection of lung cancer, with a positive likelihood ratio of 26.84 (12.60–57.20) and a negative likelihood ratio of 0.28 (0.23–0.36). Significant interstudy heterogeneity for sensitivity was observed and was likely dependent on the prevalence of malignancy in the study population and lesion size.
An overall utility of any guiding method during FOB in a work up of a pulmonary nodule was assessed in another recent meta-analysis. Authors have concluded that a diagnostic yield of guided bronchoscopic techniques is better than the traditional transbronchial approach. Although the yield remains lower than TTNA, the procedural risk appears to be lower.
Most recently another method for guided biopsy has emerged. Virtual bronchosocopy (VB) is a method of guidance where CT scanning is used to make a map of the bronchial tree, which the physician can then view on a screen while navigating the real bronchial tree with an ultra thin bronchoscope. VB differs from electromagnetic navigation in that it does not rely on real-time positional signaling from sensors around the patient. Ishida et al. used several medical centers in Japan to randomize 199 patients with peripheral lesions <3 cm to undergo a biopsy using either VB-assisted endobronchial ultrasound (VB + EBUS) or EBUS alone. A pathological diagnosis was achieved in 80% of patients in the VB + EBUS group, but only 67% of those in the EBUS alone group (p = 0.032). This is the first study that directly and prospectively evaluated the value of VB as an adjunct to EBUS-guided biopsy. Given that the evidence for VB and VB + EBUS is not nearly as robust as the evidence available for EBUS alone and/or combination of EBUS and electromagnetic navigation, it would be prudent to limit the use of this technology until more data is available.
TTNA or core biopsy, best performed under CT guidance is another useful procedure to diagnose indeterminate solitary pulmonary nodules, particularly with peripheral lesions, and can be performed on nodules less than 2 cm in size. The diagnostic accuracy of this procedure exceeds 95% for malignant lesions with slightly poorer performance for benign lesions. The diagnostic yield of TTNA can be improved even further when core biopsy is done and the cytopathologist is available to determine the adequacy of the specimen. This procedure does carry risks, including hemorrhage and pneumothorax. Pneumothorax is a much more frequent complication occurring in 10–35% of cases with only 5–10% of those patients requiring chest tube placement. In a more recent, population-based cross-sectional analysis of over 15,000 patients who underwent TTNA, hemorrhage was reported in 1% and pneumothorax in 15% with 6.6% requiring chest tube placement. Significant risk factors for those patients with complications appeared to be between 60 and 69 years of age, current smokers and those with a diagnosis of chronic obstructive pulmonary disease. Therefore, TTNA should be considered with caution in patients with poor lung reserve and postpneumonectomy.
Use of ultrasound for guidance of a needle, especially in peripheral nodules, offers a radiation-free alternative to CT-guided biopsy. The diagnostic accuracy appears to be comparable with a CT-guided approach with 92% for malignant lesions, a negative predictive value of 93% and an overall complication rate of 10% in some studies. While this approach is feasible and has been proven effective even for lesions less than 1 cm in size, great care must be taken to select appropriate patients and lesions, as well as assurance of appropriate experience of the operator.
