Ann Thorac Surg 2003;76:867-871
© 2003 The Society of Thoracic Surgeons
Original article: general thoracic
Differentiating between atypical adenomatous hyperplasia and bronchioloalveolar carcinoma using the computed tomography number histogram
Hiroaki Nomori, MD, PhDa*,
Takashi Ohtsuka, MDa,
Tsuguo Naruke, MD, PhDa,
Keiichi Suemasu, MD, PhDa
a Department of Thoracic Surgery, Saiseikai Central Hospital, Tokyo, Japan
Accepted for publication April 17, 2003.
* Address reprint requests to Dr Nomori, Department of Thoracic Surgery, Saiseikai Central Hospital, 1-4-17 Mita, Minato-ku, Tokyo 108-0073, Japan
e-mail: hnomori{at}qk9.so-net.ne.jp
 |
Abstract
|
|---|
BACKGROUND: Both atypical adenomatous hyperplasia (AAH) and bronchioloalveolar carcinoma (BAC) appear as ground glass opacity (GGO) lesions by computed tomography (CT) and are sometimes difficult to differentiate. To aid distinction between the two, we examined their CT number histograms.
METHODS: Histograms of pixel CT numbers were made for AAH (n = 9) and nonmucinous BAC (n = 8), and the peak and mean CT numbers on the histogram were quantified.
RESULTS: Although there was no significant difference in lesion size between AAH and BAC, all AAHs were less than or equal to 1 cm in diameter. All AAHs and BACs manifested one histogram peak. Both the peak and mean CT numbers on the histogram were significantly lower for AAH than for BAC (p < 0.001). However, the degree of overlap between AAH and BAC was less for the peak CT number than for the mean CT number.
CONCLUSIONS: The peak CT number on the histogram can help the radiologic differentiation between AAH and BAC. GGO lesions less than or equal to 1 cm in diameter that are diagnosed as AAH from the CT number histogram can be safely followed by CT.
|
Introduction
|
|---|
Recent advances in helical computed tomography (CT) have increased the resectability of small lung nodules and allow images of nodules to be investigated in greater detail. However, surgical resection of small nodules is sometimes needed in order to differentiate lung cancer from nonmalignant lesions [1–3]. In particular, for lesions with focal ground-glass opacity (GGO), which appear as faint nodules on CT, a number of differential diagnoses are possible, including inflammatory disease, focal fibrosis, atypical adenomatous hyperplasia (AAH), and adenocarcinoma, especially bronchioloalveolar carcinoma (BAC) [4]. Although inflammatory lesions may resolve spontaneously, or after treatment with antibiotics, both AAHs and BACs often appear as similar GGO lesions and may not enlarge for several months or more [5], making differential diagnosis difficult even at follow-up. While AAH is thought to be a precursor or even an early-stage lesion of BAC or adenocarcinoma [6–11], there have been no reports of progression of AAH to BAC or adenocarcinoma. Therefore, although BACs need to be resected surgically, lesions suspected to be AAH can be safely followed by CT.
While the term GGO is used to describe a hazy increased attenuation of the lung on CT with preservation of the bronchial and vascular margins, the criteria for GGO are as "hazy" as its image on CT scans. Several authors have classified focal GGO lesions into pure GGO and mixed GGO with a solid component, and AAH is more frequently the pure GGO lesion than BAC [5,12]. However, differentiation between pure and mixed GGO lesions is also "hazy." Therefore, to differentiate AAH from BAC objectively on the basis of CT findings, we examined the difference between the CT number histograms of the two lesions.
|
Material and methods
|
|---|
From January 1997 to September 2002, 303 patients with primary lung cancer underwent surgical resection at our hospital. Of these, 215 patients were diagnosed as having lung adenocarcinoma, of whom 12 had BAC. During the same period, 8 patients with AAH that appeared as GGO lesions on CT underwent thoracoscopic biopsy.
The histologic criteria of AAH and BAC were based on the 1999 World Health Organization histologic classification [13,14]. BAC was defined as a tumor exhibiting a pure bronchioloalveolar growth pattern with an increase in thickness of the alveolar septa and no evidence of stromal, vascular, or pleural invasion. To check for vascular and pleural invasion, we routinely conducted elastica van Gieson staining. BAC was further classified into mucinous and nonmucinous types. AAH was defined as a proliferation of minimally atypical cuboidal type II pneumocytes with gaps between the cells. The lesion was usually localized, with well defined boundaries. The atypical epithelial cells proliferated along the slightly thickened alveolar wall and had variable degrees of nuclear atypia, although the atypia was less prominent than in BAC.
The following acquisition variables were used on the CT scanner (ProSeed SA, General Electrical Medical System, Milwaukee, WI): high voltage (120 kV), tube load 160 mA, window level 500 Hounsfield units (HU), window width –1500 HU, and a 512 x 512 matrix corresponding to a pixel size of about 0.6 mm. Air calibration was conducted every morning before using the phantom. All of the lesions were subjected to helical scanning using sections 1-mm to 3-mm thick during one breath hold with maximum inspiration. Each tumor was scanned with at least three slices. A histogram was created in accordance with the menu of the CT operators’ manual. First, the tumor slice with the greatest area was selected, and the circumference of the tumor was traced on the screen. The CT numbers of each pixel within the tumor were then measured. The mean and standard deviation (SD) of pixel CT numbers within the tumor were counted, and a histogram was created. The CT number at the peak of the histogram was quantified.
All data were expressed as mean ± standard deviation (SD). The differences of mean and SD values were analyzed between the groups with using a two-tailed Student’s t test. Between-group differences at p less than 0.05 were regarded as significant.
|
Results
|
|---|
Of the 12 patients with BAC, 10 had nonmucinous BAC and 2 had mucinous BAC. We excluded the 2 patients with mucinous BAC because their CT numbers were strongly affected by mucin within the tumor, and not by the tumor growth itself. We also excluded 2 patients with nonmucinous BAC, who had another primary lung adenocarcinoma to rule out the possibility of pulmonary metastasis, which left a final total of 8 studied patients with nonmucinous BAC. All of the 8 BAC lesions were revealed as GGO lesion on CT. One of 8 patients with AAH had two lesions, so a total of 9 AAH lesions were examined. There were no significant differences in age or sex ratio between the AAH and BAC patients (Table 1). The mean values of maximum tumor diameter were 0.7 ± 0.3 cm (range: 0.3 to 1.0 cm) for AAH and 1.1 ± 0.5 (0.5 to 2.0) for nonmucinous BAC. There was no significant difference in tumor diameter between AAH and BAC (p = 0.08), but all of the AAH lesions were less than or equal to 1.0 cm. Three of eight BAC lesions were 1.3, 1.6, and 2.0 cm in diameter, respectively, whereas the other five lesisons were less than or equal to 1.0 cm. Among the BAC patients, the pathologic tumor stage was T1N0M0 in 6 patients, although the N stages of the remaining 2 patients could not be classified because of lung wedge resection.
The CT number histograms revealed characteristic features for each histologic type. The vertical line in the histograms illustrated the numbers of the pixels within the tumors. The horizontal axis revealed the CT number, for each of which the class interval was 10 HU and the range –900 to 300 HU. The histogram of AAH usually characterized one sharp peak at about –700 HU (Figure 1).
The histogram of nonmucinous BAC usually revealed one peak at about –600 HU, and the peak was more gradual than that of AAH (Figure 2).
View larger version (40K):
[in this window]
[in a new window]
|
Fig 1. Atypical adenomatous hyperplasia. (A) A nodule, 7-mm diameter, is indicated by an arrow. (B) Computed tomography number histogram, of which peak, mean, and standard deviation values are -730, -711, and 80 Hounsfield units. (A = the area of the nodule; M = the mean value of computed tomography numbers; SD = the standard deviation.)
|
|
View larger version (43K):
[in this window]
[in a new window]
|
Fig 2. Nonmucinous bronchioloalveolar carcinoma. (A) A nodule, 7-mm diameter is indicated by an arrow. (B) Computed tomography number histogram, for which peak, mean, and standard deviation values are –570, –555, and 95 Hounsfield units. (A = the area of the nodule; M = the mean value of computed tomography numbers; SD = the standard deviation.)
|
|
Figure 3
illustrates the distributions of peak CT numbers in AAH and BAC. The mean values of the peak CT numbers were –707 ± 36 HU (range: –740 to –650) for AAH and –573 ± 72 (–650 to –450) for BAC. AAH had a significantly lower peak CT number than BAC (p < 0.001). Although 1 patient each with AAH and BAC had the same peak CT number at –650 HU, there was no further overlap between the two groups. The AAH with the peak CT number at –650 HU was 1.0 cm in diameter, which was the largest in the AAH group. The BAC with the peak CT number at –650 HU was 0.5 cm in diameter, which was the smallest in the BAC group. However, there was no difficulty in histologic differentiation between these BAC and AAH lesions.
View larger version (10K):
[in this window]
[in a new window]
|
Fig 3. Distributions of peak CT number for AAH and nonmucinous BAC. (AAH = atypical adenomatous hyperplasia; BAC = bronchioloalveolar carcinoma; CT = computed tomography; HU = Hounsfield units.)
|
|
Figure 4
illustrates the distributions of the mean CT number for AAH and BAC. The overall mean value of mean CT numbers was –697 ± 56 HU (range: –749 to –566) for AAH and –541 ± 73 (–651 to –447) for BAC. AAH had a significantly lower mean CT number than BAC (p = 0.0007). The overall mean value of mean CT numbers in the five BAC lesions less than or equal 1.0 cm was –572 ± 63 (–651 to –497), which was also significantly higher than that of AAH (p = 0.01). Three patients in the two groups exhibited overlapping, ie, one AAH lesion had a mean of –566 HU and two BAC lesions had means of –651 and –622 HU, respectively. The former AAH lesion was 1.0 cm in diameter, which was the largest one in the AAH group, whereas one of the latter two BAC lesions was 0.5 cm in diameter, which was the smallest one in the BAC group.
View larger version (10K):
[in this window]
[in a new window]
|
Fig 4. Distributions of mean CT number for AAH and nonmucinous BAC. (AAH = atypical adenomatous hyperplasia; BAC = bronchioloalveolar carcinoma; CT = computed tomography; HU = Hounsfield units.)
|
|
The mean of the SD of CT number was 61 ± 43 (range: 30 to 168) for AAH, which was significantly lower than the value of 112 ± 50 (74 to 220) for BAC (p = 0.009). The mean of the SD of CT number in the five BAC lesions less than or equal to 1.0 cm was 84 ± 17 (68 to 109), which was also significantly higher than that of AAH (p = 0.01).
|
Comment
|
|---|
We examined not only the mean CT number but also the peak CT number on the histogram to demonstrate the difference between AAH and BAC for the following reasons: (1) whereas the mean CT number is easy to calculate in any CT model, it can be affected by the densities of vessels or bronchioles within the tumor; (2) the peak CT number, on the other hand, is the most frequently used value for the tumor, and allows the effect of vessels and bronchioles within GGO lesions to be ignored; and (3) GGO lesions mixed with a solid component should be represented by the highest peak number rather than by the mean value. As a result, both the peak CT and mean CT number were able to clearly differentiate between AAH and BAC. However, the peak CT number indicated less overlap between AAH and BAC than the mean CT number.
Although CT slices in the present study were ranged from 1-mm to 3-mm thick, we believe that CT numbers would not differ according to the thickness, especially in AAHs and BACs, because all of AAH and BAC in the present study were pure GGO lesions, which revealed one peak at the low CT number on the histogram, the structures of tumor, ie, densities of air spaces and cellular components, should be almost homogenous. Therefore, CT numbers should differ little according to the slice thickness, if the slices did not contain outside the tumor.
While AAH is thought to be a precursor or even an early-stage lesion of BAC or adenocarcinoma [6–11], there has been no direct evidence that AAH cells develop into adenocarcinoma cells. Therefore, AAH does not need to be resected and can simply be followed by CT. However, the differentiation of AAH from BAC by CT is often difficult. Nakata and colleagues [12] reported the following differences in CT findings between AAH and BAC or adenocarcinoma: (1) AAH lesions are usually less than or equal to 1 cm in diameter, whereas BACs and adenocarcinomas are sometimes greater than or equal to 1 cm; (2) AAH lesions often appear as pure GGO, whereas some BACs and adenocarcinomas appear as GGO mixed with solid components. However, their conclusions would not allow distinction of AAH from BACs or adenocarcinomas presenting as pure GGO and with diameters less than or equal to 1 cm. The present study was able to demonstrate distinct differences in both the peak and mean CT number between AAH and BAC, even for lesions less than or equal to 1 cm. Especially with the peak CT number, only one patient with each of AAH and BAC had the same peak CT number at –650 HU, respectively, which could represent a borderline value between the two groups. All of the AAH lesions were less than or equal to 1 cm in diameter. The SD of CT numbers was significantly lower for AAH than for BAC, indicating that the peak on the histogram was usually sharper in the former than in the latter. Therefore, we could say that GGO lesions with a diameter of less than or equal to 1 cm and a histogram CT number exhibiting one sharp peak at less than –650 HU are likely to be AAH rather than BAC.
What is the reason for the difference in CT numbers between AAH and BAC? AAH and BAC usually appear as GGO lesions on CT due to the combined effects of reduction of alveolar air spaces and increased cellular components with alveolar cuboidal cell hyperplasia, thickening of alveolar septa, and partial filling of the alveolar air spaces by tumor cells [15]. AAH usually has more air spaces and fewer cellular components than BAC or adenocarcinoma, so that the former usually has significantly lower CT number than the latter.
In conclusion, AAH can be distinguished from BAC or adenocarcinoma on the basis of the CT number histogram. Lung nodules that are less than or equal to 1 cm in diameter and have one sharp peak at less than –650 HU on the CT number histogram can be followed by CT rather than surgical biopsy. If a nodule is subsequently discovered to be a BAC, follow-up with CT would not miss the chance of surgical cure, because it would remain within the T1N0M0 stage during the follow-up period until it grew to 2 cm in diameter [16–18]. For how long does this kind of GGO lesion need to be followed by CT? Aoki and coworkers [5] reported that the mean tumor doubling time of BAC was 880 days. Therefore, a careful follow-up by CT for 2 or 3 years would be enough to confirm whether or not the lesions are AAH. However, no study has yet confirmed whether AAH lesions remain as they are, or progress to BAC or adenocarcinoma. Therefore, at the present time, it is necessary to continue following up GGO lesions that are diagnosed as AAH from the CT number histogram.
|
References
|
|---|
- Nomori H., Horio H., Fuyuno G., Kobayashi R., Morinaga S., Suemasu K. Lung adenocarcinomas diagnosed by open lung or thoracoscopic vs. bronchoscopic biopsy. Chest 1998;114:40-44.[Abstract/Free Full Text]
- Nomori H., Horio H. Colored collagen is a long-lasting point marker for small pulmonary nodules in thoracoscopic operations. Ann Thorac Surg 1996;61:1070-1073.[Abstract/Free Full Text]
- Nomori H., Horio H., Naruke T., Suemasu K. Fluoroscopy-assisted thoracoscopic resection of lung nodules marked with lipiodol. Ann Thorac Surg 2002;74:170-173.[Abstract/Free Full Text]
- Collins J., Stern E.J. Ground-glass opacity at CT: the ABCs. Am J Roentgenol 1997;169:355-367.[Free Full Text]
- Aoki T., Nakata H., Watanabe H., et al. Evolution of peripheral lung adenocarcinomas: CT findings correlated with histology and tumor doubling time. Am J Roentgenol 2000;174:763-768.[Abstract/Free Full Text]
- Kitamura H., Kameda Y., Ito T., Hayashi H. Atypical adenomatous hyperplasia of the lung. Am J Clin Pathol 1999;111:610-622.[Medline]
- Noguchi M., Shimosato Y. The development and progression of adenocarcinoma of the lung. Cancer Treatment Res 1995;72:131-142.[Medline]
- Kitamura H., Kameda Y., Nakamura N., et al. Atypical adenomatous hyperplasia and bronchioloalveolar lung carcinoma: analysis by morphometry and the expressions of p53 and carcinoembryonic antigen. Am J Surg Pathol 1996;20:553-562.[Medline]
- Nakayama H., Noguchi M., Tsuchiya R., Kodama T., Shimosato Y. Clonal growth of atypical adenomatous hyperplasia of the lung: cytofluorometric analysis of nuclear DNA content. Mod Pathol 1990;3:314-320.[Medline]
- Pueblitz S., Hieger L.R. Expression of p53 and CEA in atypical adenomatous hyperplasia of the lung. Am J Surg Pathol 1997;21:867-868.[Medline]
- Nomori H., Horio H., Naruke T., Suemasu K., Morinaga S., Noguchi M. A case of multiple atypical adenomatous hyperplasia of the lung detected by computed tomography. Jpn J Clin Oncol 2001;31:514-516.[Abstract/Free Full Text]
- Nakata M., Saeki H., Takata I., et al. Focal ground-glass opacity detected by low-dose helical CT. Chest 2002;121:1464-1467.[Abstract/Free Full Text]
- Travis W.D., Colby T.V., Corrin B., Shimosato Y., Brambilla E. World Health Organization international histological classification of tumours. Histological typing of lung and pleural tumours. Berlin: Springer-Verlag, 1999.
- Brambilla E., Tarvis W.D., Colby T.V., Corrin B., Shimosato Y. The new World Health Organization classification of lung tumours. Eur Resp J 2002;18:1059-1068.
- Kushihashi T., Munechika H., Ri K., et al. Bronchioloalveolar adenoma of the lung: CT-pathologic correlation. Radiology 1994;193:789-793.[Abstract/Free Full Text]
- Noguchi M., Morikawa A., Kawasaki M., et al. Small adenocarcinoma of the lung. Histologic characteristics and prognosis. Cancer 1995;75:2844-2852.[Medline]
- Higashiyama M., Kodama K., Yokouchi H., et al. Prognostic value of bronchiolo-alveolar carcinoma component of small lung adenocarcinoma. Ann Thorac Surg 1999;68:2069-2073.[Abstract/Free Full Text]
- Matsuguma H., Yokoi K., Anraku M., et al. Proportion of ground-glass opacity on high-resolution computed tomography in clinical T1N0M0 adenocarcinoma of the lung: a predictor of lymph node metastasis. J Thorac Cardiovasc Surg 2002;124:278-284.[Abstract/Free Full Text]
Top
Abstract
Introduction
