Between November 1994 and October 2006, we reviewed all surgical biopsy files and selected 16 patients with a pathologic diagnosis of tracheobronchopulmonary leiomyoma. Two patients with small leiomyomas (< 2 mm in maximal diameter) were excluded. The tumors in these cases were confirmed as incidental findings at bronchoscopic biopsy (bronchoscopy performed for reasons other than airway tumor evaluation, i.e., evaluation of interstitial lung disease and tuberculosis), but volume (helical) CT data were not available; only unenhanced thin-section CT scans were obtained at 10-to 20-mm intervals. Also excluded was one patient with intrapulmonary leiomyoma. Thus we included a total of 13 patients with histopathologically proven airway leiomyoma.

Cigarette smoking history and the presence of symptoms and signs were assessed. We also reviewed how the leiomyomas were detected and treatments administered in each case. Follow-up details on patients who underwent surgical or bronchoscopic tumor removal were recorded.

Initial chest radiographs and CT scans were available for all patients and were obtained within 30 days of each other (mean, 6 days; median, 1 day; range, 0-30 days). Chest radiographs were obtained in standard posteroanterior projection with high-voltage technique (125 kVp, 5 mAs, 10-12:1 grid ratio).

CT scans were obtained with helical technique: single-detector CT (HiSpeed Advantage, GE Healthcare) for five patients, 4-MDCT (LightSpeed QX/i, GE Healthcare) for one patient, 8-MDCT (LightSpeed Ultra, GE Healthcare) for two patients, and 16-MDCT (LightSpeed 16, GE Healthcare) for five patients. Scanning was performed from the level of the thoracic inlet to the midlevel of the kidney. For two patients, only unenhanced scans were obtained; for one patient, only enhanced scans; and for 10 patients, both enhanced and unenhanced scans. The 10 patients included the two with leiomyoma < 2 mm in diameter, neither of which was identified. For enhancement studies, 100 mL of contrast medium (iomeprol, Iomeron 300, Bracco) was administered IV. Imaging was started 20 seconds after completion of the contrast injection. For all patients, the scanning parameters used were 120 kVp and 170-200 mA. For single-detector helical CT, 7-mm collimation, pitch of 1, and a reconstruction thickness of 7 mm were used for scanning and image reconstruction. For 4- to 16-MDCT, a beam width of 10 mm, beam pitch of 1.375-1.5, and reconstruction thickness of 2.5-5.0 mm were used. All image data were reconstructed with a bone algorithm. Data were entered directly into a PACS (PathSpeed or Centricity 2.0, GE Healthcare), which displayed all image data on monitors (four monitors, 1,536 × 2,048 image matrices, 8-bit viewable gray scale, 60-foot-lambert luminescence). These monitors were used to view both mediastinal (width, 400 H; level, 20 H) and lung (width, 1,500 H; level, -700 H) window images.

The findings of chest radiographic and CT studies were analyzed retrospectively and jointly by two chest radiologists, who had 2 and 16 years of experience interpreting chest CT scans. Decisions on findings were reached by consensus. Analysis of chest radiographs included locating the main tumors and identifying the presence of postobstructive pneumonia or atelectasis. When these conditions were absent, lesions were regarded as solitary nodules if the diameter was less than 3 cm and as masses if the diameter was more than 3 cm.

CT analysis included determination of the location of the tumor in the airway and of the shape, size, homogeneity, and attenuation coefficients of the tumor. Tumors were subcategorized by location as tracheal or as being located in a main, lobar, segmental, or subsegmental bronchus. The tumors also were subcategorized as round, oval, or lobulated. Short- and long-axis tumor diameters were measured. Tumors were subcategorized according to internal content as homogeneous or heterogeneous on unenhanced and enhanced scans. When both unenhanced and enhanced scans were available, enhancement attenuation values of tumor nodules were calculated. The presence of postobstructive pneumonia, bronchial dilatation, mucoid impaction, and atelectasis was recorded, as were the presence and type of intratumoral calcification.

A lung pathologist with 14 years of experience reviewed all pathologic specimens. To diagnose leiomyoma, immunostaining, that is, smooth muscle actin and desmin staining, was conducted to differentiate this tumor from other spindle cell tumors, such as fibroma, neurofibroma, and schwannoma [2]. In the five patients who underwent surgical resection, macroscopic tumor appearance and relation between tumor and airways were described. During histopathologic examinations, histologic tumor grade (low or high) and the presence of intratumoral calcification, tumor necrosis, and extraluminal tumor components were evaluated.

Demographic data and imaging findings are summarized in Table 1. The patients were eight men and five women with an age range of 17-70 years (mean age, 48 years; median, 51 years). At bronchoscopy the tumors were located in the trachea (n = 4, 31%) (Figs. 1A, 1B, 1C, 1D, 1E, 1F and 2A, 2B, 2C, 2D) and in a main (n = 4, 31%) (Figs. 3A, 3B, 3C, 3D and 4A, 4B, 4C, 4D), lobar (n = 3, 23%, including two lesions in the bronchus intermedius), or segmental (n = 2, 15%) bronchus. Lesions detected with bronchoscopy and CT had the greatest diameters, ranging from 2 to 40 mm (mean ± SD, 15 ± 9.3 mm; median, 22 mm).

In two (15%) of the patients, leiomyomas were detected incidentally during bronchoscopic evaluation of the airways for other conditions (tuberculosis in one patient, lung cancer in the other). Two other patients presented with radiographic abnormalities without specific symptoms or signs. The other nine patients had symptoms including cough (n = 7), dyspnea (n = 5), sputum (n = 4), fever (n = 2), wheezing (n = 1), and chest discomfort (n =1). Six men were smokers (mean, 36 pack-years; range, 3-80 pack-years); two men and five women were nonsmokers.

Treatments administered for leiomyoma were bronchoscopic tumor resection in eight cases, tracheal segmental resection in three cases, lobectomy in one case, and segmental resection of the trachea and main bronchi with carinal reconstruction in one case. During follow-up lasting a mean of 844 days (range, 23-2,739 days; median, 379 days), CT (n = 7) and bronchoscopy (n =1) showed no evidence of recurrence. One patient (patient 7) who underwent surgical tumor removal had undergone bronchoscopic removal of an endobronchial leiomyoma approximately 18 months previously.

Chest radiographic findings were normal in three patients with a small tumor nodule. In the other 10 patients, the tumor was identifiable and appeared as an endobronchial nodule with postobstructive pneumonia or atelectasis (n = 5) (Fig. 4A, 4B, 4C, 4D), an endotracheal nodule (n = 3) (Fig. 1A, 1B, 1C, 1D, 1E, 1F), or an intratracheal and extratracheal (n = 1) (Fig. 2A, 2B, 2C, 2D) or intrabronchial and extrabronchial (n = 1) (Fig. 3A, 3B, 3C, 3D) mass.

Lesions were identified in 11 patients. In two other patients (patients 8 and 12), lesions were small and were found incidentally during bronchoscopy performed to evaluate other lung lesions. CT images reconstructed with a 7-mm section thickness did not depict these lesions. The lesions manifested as an airway intraluminal nodule in nine cases (Figs. 1A, 1B, 1C, 1D, 1E, 1F and 4A, 4B, 4C, 4D) and as an iceberg tumor (small intraluminal component and large extraluminal component) in two cases (Figs. 2A, 2B, 2C, 2D and 3A, 3B, 3C, 3D). Lesions had an oval contour in seven cases, a lobulated contour in three cases, and a round contour in one case. Obstructive pneumonia, atelectasis, or mucus plugging was found in five patients in whom a tumor nodule was located in a main (n = 2), lobar (n =2), or segmental (n = 1) bronchus. The internal contents of tumors were homogeneous in nine of the 10 patients for whom unenhanced CT scans were available. In the tenth patient, the tumor was heterogeneous with a greatest diameter of 22 mm. On the enhanced scans available for nine patients, five tumors (greatest diameter, 25 mm; mean, 20 ± 6.9 mm; median, 23 mm) had homogeneous enhancement, and four had heterogeneous enhancement (greatest nodule diameter, 40 mm; mean, 26 ± 17.3 mm; median, 20 mm). In eight patients for whom both unenhanced and enhanced scans were available, one lesion had a heterogeneous appearance on unenhanced scans, and three lesions, including the lesions with a heterogeneous appearance on unenhanced scans, had a heterogeneous appearance on enhanced scans (Table 1). Leiomyomas had an attenuation of 25-46 H on unenhanced CT and 46-85 H on enhanced CT (mean attenuation, 32 ± 14.7 H; range, 15-60 H; median, 30 H). Stippled calcifications were found in a patient (patient 2) with a large tumor that had both intraluminal and extraluminal components.

In the five patients who underwent surgical resection, CT precisely depicted tumor location and extent. The macroscopic appearances of these tumors were whitish-tan and oval or lobulated margin. The tumors were polypoid intraluminal masses in three patients and intraluminal and extraluminal masses in two patients. Microscopic examination showed all tumors were composed of proliferating spindle cells. Immunostaining for smooth muscle actin and desmin was performed in four cases, and all results were strongly positive. No cellular pleomorphism was found. Mitotic figures were found in one patient but were not prominent (1/20 high-power fields). No tumor necrosis was found. Intratumoral dystrophic calcification was found in one patient. Histologic results revealed reactive hyperplasia in the regional lymph nodes of two patients.