Abstract
OBJECTIVE To determine the possibility of definitive diagnosis for solitary pulmonary nodules in patients with a primary extrathoracic malignant neoplasm (ETM-SPN), and to further evaluate the value of CT for differential diagnosis in ETM-SPN by a multivariate retrospective study.
METHODS Eighty-three patients with pathologically and clinically proven ETM-SPN with a diameter smaller than 3 cm were included in this study. The pathological characteristics of the SPN were correlated with those of the extrathoracic neoplasm, with the patient's age, gender, smoking history, disease-free time interval between the diagnosis of the extrathoracic malignancy and that of the lung lesion. In all 83 cases, CT scans were reviewed to confirm the solitary nature, size, and nodular morphology of the lung lesion.
RESULTS Of all 83 cases, the mean age was (57.43±15.34) years. There were 51 males and 32 females, with the ratio of 1.59:1. The lesions included solitary metastasis in 43 cases, pulmonary malignant lesions in 33, and benign lesions in seven. Between the primary lung cancers and solitary metastasis groups, there was no significant difference in the gender ratio (1.20:1 vs 2.31 = 1, X2=0.0209, P>0.05), but there was a significant difference between the mean age (62.48 ±11.96 years vs 54.10±16.49 years, t=3.34, P<0.05). In the primary lung cancer and metastasis patient group, the percentage of patients who had a smoking history were 39.3 %(11/17) and 35.9 %(14/39), respectively. Patients with a primary lung cancer had no significant higher frequency of smoking history than did those with a metastatic lesion (X2=0.640, P>0.05). Of 81 cases who were followed-up, the mean time of the disease-free interval between extrapulmonary malignancy diagnosis and pulmonary lesion differentiation was 39.73 ± 6.29 months (range 0~300 months, median 20.00 months), whereas those in the primary lung cancer group and metastatic group were 65.62 ±13.45 months and 22.83 ±4.19 months respectively. This difference was significant between the two groups (Wilcoxon rank sum test, U=2.796, P<0.01). Of all 83 cases, there were ten extrapulmonary squamous carcinomas and 58 adenocarcinomas with ratio of primary lung cancer and solitary metastasis of the tumors were 7: 3 and 24:34, respectively (χ2 =1.781, P >0.05), without showing a statistically significant relevance between the pathologic patterns of extrapulmonary malignancy and characteristics of the lung nodules. Of all the 83 cases, the mean diameters were (2.77 ±1.25) cm, whereas the diameters of 33 cases of primary lung cancer and 43 cases of a solitary metastatic lesion were (2.86±1.18) cm and (2.62±1.31) cm, respectively. There was no association between the two groups (t =1.29, P>0.05). There was a statistically significant association between primary lung cancer and the metastatic group with spiculate and smooth edges of the lung lesion (X2=8.562, P<0.01; χ2=15.220, P<0.001). The study showed that a lung nodule with a spiculated margin correlated with a primary lung carcinoma, whereas those nodules with a smooth edge may more frequently show as a metastastic pulmonary lesion. CT-pathologic correlative analyses of hilar and mediastinal adenopathy were reviewed in 37 patients who underwent lobectomy and thoracotomy. There was no statistical significant difference between the primary lung cancer group and the metastatic group (χ2=2.801,P>0.05).
CONCLUSION The likelihood of a primary lung cancer versus a metastasis of ETM-SPN smaller than 3 cm mainly depends on the patient's age, free interval between the two tumors and CT morphological characteristics of the lung lesion. This study showed there was no significant relevancy to factors such as gender, smoking history, pathological patterns of the extrapulmonary neoplasm or whether there has hilar or mediastinal adenopathy.
keywords
It is not rare for a patient with an extrathoracic malignancy to develop a solitary pulmonary nodule (solitary pulmonary nodule in a patient with an extrathoracic malignancy, ETM-SPN), which tends to be primary lung cancer, a solitary metastasis or benign lesion. The histology of such a nodule is highly important for the determination of staging classification, the selection of appropriate therapy and the evaluation of prognosis. We are in urgent need of a feasible diagnostic method. Some scholars have analyzed the correlation between a pulmonary mass and primary extrapulmonary tumors. However, many questions need to solved, such as the importance of the size of the lesion, the time interval between the two tumors and the correlation between morphological characteristics on CT and the histology of the pulmonary nodule.
The authors of this report retrospectively analyzed 83 cases with the hope of finding a method to aid in the determination of the pathology of the SPNs in patients with an extrapulmonary malignancy. In this report we have discussed the probability of a patient having a primary lung cancer versus a benign lesion and the presence of metastases.
Materials and Methods
Clinical data
We retrospectively evaluated SPN cases, which were examined by CT, with a history of primary extrathoracic malignancy between January 1992 and April 2003 in the Changzheng hospital. There were totally 83 cases, 51 males and 32 females; ages ranged from 12 to 85 years with a median of 60 years. All the SPNs were detected after or at the time of diagnoses of the extrathoracic malignancies, all of which were primary rather than secondary. The patient’s age was that at the time of diagnosis of the pulmonary tumor, or at the first detection by a CT scan for the patients during follow-up. The monthly tumor-free time interval was determined by the time of pathological diagnosis. Patients who received no operation but who had been treated (e.g., chemotherapy, radiation therapy) were excluded. Lesions which showed no change or reduced in size during the 24 months of following-up were considered to be benign.
Equipment for examination
CT scans used in this study included SIEMENS Somatom CR, AR,Plus S and PHILIPS Mx8000 multislice scanner, with a spiral CT scan as the major technique. The slice thicknesses mainly used included 1.0, 1.6, 2.0, 2.5, 3.2, 5.0 and 10.0 mm. The mediastinal window (level 50 Hu,width 300~400 HU) was selected to measure and estimate the maximum diameter, lobulation, calcification, and lymph nodes in the mediastinum and hilus of the lungs. The lung window (level 50 Hu, width 300~400 HU) was selected to depict morphological characteristics such as spiculation and the tumor-lung interface. CT-pathological correlative analyses were performed. All the cases were reviewed by two experienced chest radiologist.
The means and methods for pathological confirmation
All the cases, except two inflammatory lesions confirmed by clinical practice, were verified by pathology (33 cases of pulmonary lobectomy, pneumonectomy and wedge excision; 24 cases of percutaneous puncture; nine cases of bronchofibroscope; seven cases of Video Assisted Thoracic Surgery (VATS); six cases of pulmonary lobectomy combined with percutaneous puncture; two cases of pulmonary lobectomy combined with bronchofibroscope). We determined the histology of pulmonary lesions through routine pathology and immunohistochemistry, and compared the results with the extrathoracic malignant neoplasms. Quint’s diagnostic criterion for primary pulmonary neoplasms and solitary metastasis was adopted when extrapulmonary and pulmonary lesions showed consistent histology.
Patient history and statistical analysis
Case number, age, gender, smoking history of the patients in the extrathoracic malignancy group and the pulmonary group, tumor-free time interval between the two tumors and the maximum diameter of pulmonary lesion were recorded. Morphological characteristics of the pulmonary lesions based on CT, such as tumor-lung interface, lobulation, calcification, lymph nodes in the hilus of the lungs and mediastinum, as well as the correlation between the extrathoracic malignancy and the pulmonary lesion were evaluated. The percentage of the smoking population was based on those who smoke more than 400 cigarettes per year. These criteria are set by the WHO. The percentage of patients with a primary lung cancer, a solitary metastasis, or a benign lesion was tabulated for each group. SPSS 10.0 software package was used for statistical analysis and the χ2 test was used to determine the statistical significance of associations between groups. T-test, U-test and Wilcoxon rank sum tests were used to compare the medians between groups. P-values of <0.05 were considered to indicate a statistically significant difference between groups.
Results
The relevance among age, gender and the pathology of ETM-SPN
We retrospectively identified 83 consecutive patients, male vs female =1.59:1, with the average age of 57.43 ± 15.33 years. Of these cases, 39.8% (33/83)were primary bronchogenic carcinomas, 51.8% (43/83) were solitary metastatic tumors and 8.4% (7/83)were benign pulmonary lesions. The average age of patients in the lung cancer group was 62.48 ± 11.96 years, (range 37~85, median 64.00) while that of patients in the solitary metastasis group was 54.10 ± 16.49 years (range 12~79,median 54.00). There was a statistically significant difference between the ages of the two groups (t=3.34, P<0.05). The gender ratio of male to female was 1.20:1 (18:15 )in the primary lung cancer group, and that in the solitary metastasis group 2.31:1 (30:13)(χ2 = 0.0209,P> 0.05).
The role smoking history plays in the pathology of ETM-SPN
The relevance between smoking history and histology of the lesion for 73 cases with definite smoking or nonsmoking history (five, four and one cases dropped out in primary lung cancer group, solitary metastasis group and benign lesion group, respectively.) is shown in Table 1.
The relevance between 73 cases of nonsmokers and smokers in the pathology of pulmonary lesions
Relevance between the interval time of the two lesions and the pathology of ETM-SPN
We identified 83 cases while two dropped out. The average interval between the detection of the two groups was 39.73 ± 6.29 months (0~300 months, median 20.0 months), while that of the 32 cases in the lung cancer group was 65.61 ± 13.45 (0~300, median 36.0)months and the 42 cases in the solitary metastasis group was 22.83 ± 4.19 months (0-156 months, median 15.0 months). There was a significant difference between the two groups (Wilcoxon rank sum test, U= 2.796, P < 0.01).
Relevance between the pathology of extrathoracic malignancy and ETM-SPN
We analyzed the relevance between the pathology of extapulmonary malignancy and ETM-SPN in cases who had extrapulmonary adenocarcinoma and squamous cell carcinoma. There was no significant difference between the lung cancer group and the solitary metastasis group, as was the case with the solitary metastasis group and the benign lesion group (Table 2).
Relevance of the pathology of extrapulmonary tumors and ETM-SPN of seventy-four cases
Relevance between morphological characteristics and the pathology of the lung lesions (primary or secondary)
The average maximum diameter of the 83 solitary pulmonary nodules was 2.77 ± 1.25 cm, while the diameters in the primary lung cancer group and solitary metastasis group were 2.86 ± 1.18 cm and 2.62± 1.31 cm, respectively. There was no significant difference between the two groups(t= 1.29, P> 0.05). The relevance between pathology of the lung lesions and morphological characteristics by CT, such as tumor-lung interface, lobulation, spiculation, smooth margin, ill-idenfied margin and calcification are shown in Table 3.
Margin and morphological characteristics of 83 lung lesions
In the metastasized group, calcification occurred in five cases (3 cases were osteosarcoma, 2 were rectal adenocarcinoma) of the solitary lesions, with an occurrence of 11.6% (5/43). Three of the calcifications were diffused and two were nodules, whereas three cases in the primary bronchogenic carcinoma group showed calcification, with an occurrence of 9.1% (3/33), two were squamous cell carcinomas, the other one was adenocarcinoma. One piece of punctual calcification was seen in two cases and two pieces of calcification were seen in one case. The pattern and quantity of calcifications were different, but because of the small case number they were not analyzed further. We retrospectively identified 37 cases that had undergone resection of lymph nodes in the hilus of the lungs and mediastinum. We found in a CT-pathologic controlled study that there was no significant difference between the primary lung cancer group and solitary metastasis tumor group in the infiltration of cancer cells in the hilus of the lungs and mediastinum (Table 4).
Adenopathy of hilus or mediastinum of 37 cases and its relationship with primary or secondary lung tumors
discussion
Among the solitary pulmonary nodules in patients with a primary extrathoracic malignant neoplasm (ETM-SPN), primary lung cancer accounts for 20~60% and solitary pulmonary metastasis tumors make up 23~46%. [1, 2] Nodules less than 3 cm in diameter, solitary pulmonary metastatic tumors account for only 51.8%. So it is of significant importance for differentiation and pathological identification.
Morphology of the pulmonary nodules combined with relevant factors such as gender and age should be taken into consideration for the differentiation of ETM-SPN. Although some authors have expressed the possibility that primary lung cancers have nothing to do with age,[3, 4] we found in our research that there was a significant difference between primary lung cancer and solitary pulmonary metastatic tumor groups in average age, with elder patients being more likely to have primary lung cancer. [1] The results showed that age is one of the prognostic factors in the evaluation of ETM-SPN. The findings also indicated that with time, tumors were more likely to reoccur than to relapse and metastasize. There was no significant difference between the primary lung cancer group and metastasized tumor group in gender, which is consistent with other authors' finding. Therefore one should not consider gender as a predict factor in the diagnosis of ETM-SPN.[1]
The findings of Quint et al.[1] showed that smoking was relevant to the pathology of primary lung cancer, the incidence of smokers' primary lung cancer was 3.5 times of that of nonsmokers in ETM-SPN. However, smoking is a common cause of pulmonary lung cancer and other carcinomas. There was no significant difference between the primary lung cancer group and metastatic tumor group in smoking rate of the 73 cases with definite follow-up results. Consequently we believe that smoking should not be considered as a factor in differentiation of these tumors.
There have been few studies on the time interval between intra- and extra-lung cancers in the differentiation of ETM-SPN. The average intervals for the primary lung cancer group and metastatic tumor groups were 65.6 and 28 months, respectively, with the former being obviously longer than the latter. (U = 2.796, P < 0.001). ETM-SPN was more likely to be a primary lung cancer than a metastatic tumor when the interval was longer than 24 months (χ2 = 4.350, P < 0.05), reflecting that with time, the incidence of metastatic tumors decreased while that of primary lung cancer increased. Although some researchers have shown that when patients have an extra pulmonary adenocarcinoma, the incidence of a solitary pulmonary metastasis tumor was higher than that of extrapulmonary squamous cell carcinoma,[5] our result showed that the pathology of an extrathoracic tumor had nothing to do with the pathology of ETM-SPN such as primary lung cancer and solitary metastatic tumor when the SPN was less than 3 cm in diameter. There may be some bias among different samples due to the proportion of various kinds of tumors. The pathology of extrathoracic tumors can not be considered as a diagnostic factor for ETM-SPN.[6]
The morphology of solitary pulmonary nodules is a dominant factor in the differentiation of ETM-SPN, according to some authors.[7] To some extent, the size of the SPN was relevant with its nature; small nodules were more likely to be benign. Nakamura et al.[4] and Henschke et al.[8] found that 60% and 67% of the nodules that could only be detected on CT and VATS were benign. Another 41-case report, including osteosarcoma, melanoma, soft-tissue sarcoma, Ewing’s sarcoma, rectal adenocarcinoma and carcinoma of testis, found that 33 percent of SPNs less than 0.5 cm in diameter were metastatic tumors,[9] while 90 % of SPNs larger than 0.5 cm in diameter were metastatic tumors. However, it was difficult to differentiate small benign nodules from small inflammatory lesions, [8] such as hamartoma, sarcoidosis, silicotic granuloma, histoplasmosis, tubercler inflammatory pseudotumor, intra pulmonary lymph node, small pulmonary infarcted lesion, focal fibrosis and interlobar fissure pleural thickening. Nodule size has limited importance in the diagnosis of a real nodule. Our result also showed that there was no significant difference in size between primary lung cancer and a solitary metastatic tumor, so the tumor-lung interface, morphology, and inner density should be considered as key points.
There have been few reports concerning the tumor-lung interface of solitary pulmonary metastatic tumors based on CT. Hatanaka [10] identified 87 nodules with 5-20 mm diameters based on a HRCT-pathology control study, finding that 38 % of metastatic nodules had a smooth edge, 16 % were clear but irregular while 30 % were blurry. In our study, six cases showed spiculation, 21 showed a smooth edge while the edge of two cases was blurry, one of which was metastatic rectal carcinoma, pathology demonstrating a blur and infiltration of malignant cells. A "halo sign" was rare in primary lung cancers.[11, 12] There was a significant difference in the spiculation of the primary lung cancer group and a smooth edge in the solitary metastatic tumor group, suggesting the likelihood of primary lung cancer and a solitary metastatic tumor, respectively. [9] It is of significant value to identify primary and secondary tumors as well as the pathology of the secondary tumor according to its morphological features on CT. The study of Nakagawa [13] showed that 88 % of thyroid carcinoma, 85% of liver carcinoma, 75% of metastatic lung tumor from the kidneys were spherical lesions with a smooth edge; while 75% of laryngopharynx carcinoma, 62% of rectal carcinoma and 58% of breast cancer were irregular in form. However, our results showed that there was no significant difference in lobulation between metastatic tumors and primary tumors. So lobulation could not be considered as a factor in differentiation. There have been no reports on a solitary metastasized nodule with a cavity of ETM-SPN, and both of our two cases were primary pulmonary carcinoma, so primary lung cancer should be considered first with the exclusion of tuberculosis.[14]
It is important to identify the presence and quantity of calcification for ETM-SPN. Calcification can be detected in benign lesions such as tuberculoma and hamartoma, malignant tumors such as osteosarcoma, rectal carcinoma, chondrosarcoma, salivary gland sarcoma, ovarian tumors, breast and thyroid malignancy, as well as primary lung cancer. [15] However, the calcification of metastatic tumors, which was mostly detected in osteosarcoma and secondary pulmonary carcinoma from the rectum, [16] differed from that of primary lung cancer as diffused and large in quantity. All the 5 cases in our study were plaque-like and diffused. Therefore, it was important to perform thin-slice scans before operation so as to identify the existence of fat or calcification,[17] i.e. so to at least contribute to the exclusion of primary lung cancer. However it often leads to misdiagnosis of benign tumors without calcification. An osteosarcoma, which was pathologically identified as hamartoma, was misdiagnosed as metastatic carcinoma for there was no fat or calcification with CT.
Although, it is the most common feature for lymphoma, other tumors such as melanoma, sarcoma and carcinoma of testis have enlarged lymph nodes in the hilus of the lung and mediastinum.[3] However, it is offen considered that it is uncommon for metastatic pulmonary nodules to have swollen lymph nodes in the hilus of the lungs and mediastinum, so the detection of swollen lymph nodes on CT often suggests the likelihood of primary lung cancer, with the exclusion of diseases indicated previously.[1] However, there was no obvious relevance between the size of the lymph nodes and the infiltration of malignant cells. It has not been a routine practice to take out lymph nodes in the hilus of the lungs and mediastinum for metastatic tumors, which differs from an operation for primary lung cancer, so there have been few CT-pathology controlled studies.[18] It was verified in our study that swollen lymph nodes on CT could have normal pathology. Loehe et al. [18] did a CT-pathology controlled study on 63 cases of lymph nodes with shorter diameter less than 10 mm, finding that at least one lymph node infiltrated by cancer cells could be detected in 14.3% of cases, most of which were considered as normal on CT. McCarville et al. [19] reported that the occurrence of mediastinal lymph nodes in benign lesions (7/12, 58%) was four times that of malignant pulmonary nodules (3/21, 14%) in an osteosarcoma group. Our CT-pathology controlled study identified 37 cases with infiltration of cancer cells in the hilus of the lungs and mediastium, showing that there was no significant difference between the pulmonary lung cancer group and metastatic tumor group, which indicated that we could not define the histology of ETM-SPN by the existence of metastatic lymph nodes in the hilus of the lungs and mediastinum. Consistent with the findings of other authors, it was not specific to identify the pathology of ETM-SPN by existence of metastastic lymph nodes in the hilus and mediastium, and specificity was even poorer based on lymph node size.
Our preliminary results showed that there was diagnostic value in identifying the nature of ETM-SPN less than 3 cm in diameter by age of the patient, lobulation and spiculation on CT, as well as the interval between the confirmation of extra- and intrapulmonary lesions. There was no significant difference in evaluating ETM-SPN by gender, smoking history, pathology of extra pulmonary carcinoma, and metastatic lymph nodes in the hilus of the lungs and mediastinum.
What should be specially pointed out is that it is still difficult to identify SPNs as primary pulmonary or metastatic tumors when intra- and extrapulmonary tumors are the same in pathology. Immunohistochemistry can help differentiation, for there are specific markers for tumors from different tissues, for example, progesterone receptor (PR) and estrogen receptor (ER) are often positive in breast cancer. [20] Besides immunohistochemistry, some related diagnostic models can help define the nature of the tumor. For example, Lefor et al. [21] identified a palpable tumor as primary or secondary by the existence of swollen lymph nodes in the anterior region of the neck. When there were palpable swollen lymph nodes, they considered the SPN as metastatic; otherwise, primary lung cancer was the first choice. However, the extirpation of jugular lymph nodes may interfere with the normal passing of the lymphatic circulation, so metastatic lesions at the beginning can cause swelling of jugular lymph nodes. As a result, patients with pulmonary squamous cell carcinoma, who have no enlarged jugular lymph nodes but have undergone extraction of lymph nodes in the anterior region of the neck, may actually be a metastatic tumor rather than primary lung cancer.[1] However, clinical practice has shown that most palpable squamous cell carcinomas patients, who have undergone extraction of jugular lymph nodes, did not develope multiple metastatic tumors, demonstrating that though there were some errors in pathological classification, it was a diagnostic model with clinical value. [21] To some extent, we should treat ETM-SPN as primary carcinoma before pathological diagnosis.
- Received November 19, 2004.
- Accepted March 24, 2005.
- Copyright © 2005 by Tianjin Medical University Cancer Institute & Hospital and Springer
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.