| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1996;111:827-832
© 1996 Mosby, Inc.
GENERAL THORACIC SURGERY |
Received for publication April 27, 1995 Revisions requested July 17, 1995; revisions received Dec. 11, 1995; Accepted for publication Dec. 14, 1995. Address for reprints: Daniela Kandioler, MD, Department of Surgery, University of Vienna, Währingergürtel 18-20, A-1090 Vienna, Austria.
Abstract
When solitary pulmonary tumors are observed in patients with a history of cancer, differentiation between metastasis and primary lung cancer is crucial for appropriate therapy. Assuming that p53 mutations are conserved in metastases, mutation analysis of the p53 gene would be a valuable tool in differentiating metastases from primary carcinomas of the lung. In nine of 267 resected lung tumors, the origin of the lung tumor could not be defined histologically. Five patients had a history of colorectal carcinoma, one had a history of breast carcinoma, one had a history of soft-tissue carcinoma, and one had a history of head and neck carcinoma. One patient with a clear cell carcinoma of the lung had been surgically treated for both renal and thyroid cancer. Material from one patient with adenocarcinoma of the lung, histologically defined regional lymph nodes, and distant brain metastasis served as a control. We extracted deoxyribonucleic acid from the snap-frozen tissue of the unclassified lung tumors, from paraffin-embedded tissue of the previously removed primary cancers, and also from peripheral blood of the patients. Exons 2 to 11 of the p53 gene were amplified in separated polymerase chain reactions and directly sequenced. In all cases, the presence of germline mutations was excluded by analysis of peripheral blood deoxyribonucleic acid. The p53 mutation detected in the deoxyribonucleic acid of the lung tumor of the control patient proved to be conserved in the lymph nodes as well as in the brain metastasis. In two cases, the lung tumors exhibited a p53 mutation not present in the previously removed primary tumor and were therefore classified as new primary lung cancers. In five cases, the lung tumors proved to be metastases of the first tumor, exhibiting the identical p53 mutation. One of these lung tumor samples could be identified as a metastasis from the renal cancer, but the corresponding thyroid cancer material was different. For two cases, molecular analysis remained inconclusive. In one case, no p53 mutation could be found in the compared samples; in the other, no deoxyribonucleic acid could be extracted. Analysis of p53 mutations allowed exact classification in tumors for which standard methods failed to distinguish between metastasis or primary tumor. More than two thirds of lung tumors in patients with previous gastrointestinal carcinoma were revealed to be metastases, but second primary lung cancer could also be diagnosed. This diagnosis allowed correct surgical and adjuvant treatment of these patients. (J THORACCARDIOVASCSURG1996;111:827-32)
Adenocarcinoma of the lung, colorectal cancer, and breast cancer are high-prevalence malignant tumors. Both colorectal cancer and breast cancer tend to metastasize to the lung. Histologically, adenocarcinoma cells from lung tumor biopsy samples may represent a primary lung tumor but may also originate from adenocarcinoma of the intestine or the breast. The incidence of second primary lung cancers in patients with a history of such malignancies is stated in some centers to be about 70%.
1 Because techniques for operation and adjuvant treatment strategies differ for primary lung cancer and metastatic disease, a reliable method for differentiation of metastases from second malignancies is important. Adenocarcinomas of the intestine and of the lung do not express characteristic surface markers that could be used for differentiation. In breast cancer, however, positive test results for progesterone or estrogen (present in about half of cases 2) are valid for diagnosis.
In most malignancies, the development and clonal expansion of a tumor are preceded by acquisition of one or more genetic aberrations, which provide tumor cells with a growth advantage. Mutations within the p53 tumor suppressor gene are the most common genetic alterations associated with malignant tumors. 3,4 The gene product, a 53 kd nuclear phosphoprotein, seems to be involved in transcription, deoxyribonucleic acide (DNA) synthesis, DNA repair, programmed cell death, and control of angiogenesis factors important for metastases. Thus p53 functions as an important checkpoint in carcinogenesis and tumor progression. 3,5 One important feature of p53 mutations is their high frequency in different tumor types, irrespective of stage or tumor histology (e.g., lung cancer 50% to 70%, colon cancer 50%, and breast cancer 20% 6). The probability of detecting a p53 mutation in one of two independent tumors from the same patient is therefore high and indicates the different origins of these tumors. Another characteristic of p53 mutations is their great variability concerning site and type of mutation. Even if both independent tumors exhibit p53 mutations, it is unlikely that these mutations will be identical (both the same type and site). Taking advantage of these unique attributes of p53 mutations, we concluded that identical mutations detected in two different tumors sites of a patient are indicative of the presence of metastatic disease. The aim of our study was to show that mutation analysis of the p53 gene provides a valuable tool for differentiation between metastases and second malignancies.
Patients and methods
Patients' data are listed in detail in Table I.For nine of 267 resected lung specimens from patients with solid lung tumors who underwent operation at the Department of Cardio-Thoracic Surgery of the University of Vienna between August 1992 and September 1994, the origin of the tumor could not be defined histologically. Lung specimens were collected and stored in liquid nitrogen after informed consent was obtained from the patients. All nine patients had a history of carcinoma of other location. The lung tumors could not be classified as metastasis from the previous cancer or primary lung cancer by histologic examination (Table I).
|
7 Results
In one control patient with adenocarcinoma of the lung, histologically defined brain metastasis, and regional lymph node metastases, a p53 mutation in exon 5, which proved to be present in the lung sample, the lymph node, and the brain metastasis, led to the generation of heteroduplex bands visible on polyacrylamide gels (Fig. 1). Direct sequencing characterized this mutation as a G-T transversion in codon 157 of exon 5 of the p53 gene.
|
Table II).Patient 42, who exhibited a single pulmonary nodule consisting of adenocarcinoma cells 2 months after resection of receptor-negative breast cancer, exhibited the wild-type sequence in the breast specimen, whereas the lung tumor showed an eight-base deletion in exon 4 (Fig. 2). Patient 65 carried a mutation in the resected rectum carcinoma, whereas the lung tumor showed wild-type sequence in the p53 gene. In five cases (38, 3, 311, 234, and 96) the primary cancer and the lung sample showed identical p53 mutations (Table II). Patient 96, with a history of resected renal and thyroid cancer, exhibited a lung metastasis from a clear cell carcinoma. No mutation could be found in the thyroid cancer, but because of identical mutations in the material from the renal carcinoma and the lung tumor, the origin of the latter could be determined (Table II). In patient 311, who had resection of a rectal cancer followed by resection of a separate colon cancer, the lung tumor could be identified as a metastasis from the rectum on the basis of identical p53 mutations in both samples. Analysis in patient 8, with a history of head and neck cancer and squamous cell carcinoma of the lung, remained inconclusive because no p53 mutation could be found.
|
|
Discussion
Cancers of similar histologic types in different organs often share similar histologic and even similar immunohistochemical features.
For patients with previous head and neck cancer and later lung tumors, Chung and associates 8 showed with p53 analysis that about two thirds of the lung tumors represent second primary cancers. For the first time, we analyzed solitary adenocarcinomatous lung tumors in patients with a history of cancer of the gastrointestinal tract or breast cancer. Lacking a tool for exact diagnosis, such tumors are often treated as metastases. However, p53 mutation analysis may provide an exact molecular genetic diagnosis in such cases. In our series, two of seven suspected lung metastases proved to be primary lung cancers. Incidentally, because of the location of the tumors these two patients had been treated with pneumonectomy and lobectomy, correct treatments for primary cancers.
The high frequency of p53 mutations in lung cancer without distant metastases of all stages, including stage I, 9 indicates that p53 mutations in lung cancer are present before metastatic progression. In breast cancer, the p53 mutation rate is about 20%. 10 In colorectal cancer, 11 the p53 mutation appears to occur during progression to malignancy at a rate of 50%.
Patients with a history of colon or breast or head and neck cancer 8 and solitary lung tumors are the main candidate groups for p53 analysis. Because of the frequency of such constellations, p53 sequencing analysis can be of important clinical value. Additionally, p53 sequencing analysis can be used for determination of poorly differentiated tumors, which often do not express any markers, and also to determine how many so-called second primary lung cancers are actually first primary lung cancers metastatic to the lung.
In future, such analyses will lead to correct calculation of the risk of second primary lung cancer. Obviously, the probability of second primary lung cancer should be estimated depending on the location of the first malignancy. The frequency seems to be higher in cases of a history of head and neck cancer 8 than among our patients with previous gastrointestinal or breast cancer. Although the risk in our group was less than one third, the possibility of second primary lung cancer in such patients should be kept in mind.
Several authors have provided much evidence that p53 mutations are the most common genetic alterations among all known cancer-related genes. 4,5 The characterization of p53 mutations can therefore be used to identify the tumor cells and their assignment to the tumor clone, taking advantage of the fact that the tumor-specific genetic mutations are conserved in metastases, as was demonstrated in several articles for the p53 gene. 8,12,13 In tumor cells, the genetic changes are responsible for knocking out cell-cycle control, which leads to uncontrolled proliferation. These mutations are therefore essential for the persistence of a tumor, and it is truly improbable that they would disappear as a result of DNA repair. Because p53 enables DNA repair by arresting the cell cycle in G1 phase, 4 inactivation by mutation should lead to an accumulation of additional mutations in the genome of the tumor cell, which might coincide with tumor progression. 11 Apparently, however, these additional mutations do not occur in the p53 gene itself; we did not observe any additional p53 mutations in any distant metastases even after a period of 10 years (patient 96).
P53 abnormalities often have been studied by immunohistochemistry, which takes advantage of the fact that some mutations in the p53 gene increase the half-life of the p53 protein, which is not detectable under normal circumstances. There are several reasons for accumulation of p53 protein, however, that do not always originate from mutations of the p53 gene. Additionally, it is well known that some distinct mutations increase the instability of the protein. 14 Moreover, p53 overexpression can occur in the absence of p53 mutation, as a cellular answer to genetic damage somewhere in the genome. 14
As a non–labor intensive method, immunohistochemistry is a practical tool to screen for p53 abnormalities. A molecular genetic differentiation of tumors, as shown in our work, depends on the possibility of discriminating tumors on the basis of different types and locations of mutations. Only sequencing, which is in fact a labor-intensive method that is not currently routinely available, provides this information. A method such as p53 sequencing analysis, however, which is valid for differentiation between metastases and second primary malignancies of the lung, can support correct estimation of prognosis and allow appropriate surgical and adjuvant approaches in these patients.
Appendix: Discussion
Dr. Jack A. Roth (Houston, Texas)
I thank the authors for providing a copy of their manuscript and congratulate Dr. Kandioler and colleagues for an interesting presentation. I think that this study is well done, and it is certainly encouraging to see other thoracic surgical groups making important translational research observations in the molecular biology of cancer. I think that this may in fact represent one of the first potential applications of this type of technology to clinical practice.
The p53 gene is important in the process of carcinogenesis. It has been characterized as the guardian of the genome. This is perhaps somewhat excessive, but it does point out that if there are mutations or deletions in the p53 gene, this prevents the cell from arresting in the G0-G1 phase of the cell cycle and repairing its DNA in response to damage. This induces genetic instability and may contribute to progression of the cell through the process of transformation.
Two years ago, we published a study that showed that primary tumors of the lung, esophagus, and head and neck had discordant p53 mutations compared with second primary tumors that arose in the same patient, whereas the same mutation, as Kandioler and colleagues showed here, was present in the primary tumor and documented metastases. This has important implications for the process of carcinogenesis. It shows that multiple primary cancers of the aerodigestive tract can arise in a field that is exposed to a carcinogen and that these cancers arise as clonally independent events.
In this study, the technique was used to distinguish between unrelated primary cancers that arise at different sites and metastases. It is encouraging that the same findings were observed as we reported in our study.
The determination of p53 mutations is now routine in many institutions, and I think it should be possible to apply these observations clinically. One reservation, of course, is that the same genetic change could be observed to occur in separate primary tumors. I think that this is a highly improbable event; however, it must be kept in mind that there are certain hot spots in the p53 gene and such a coincidence is therefore at least a theoretic possibility.
I have two questions. First, six of the eight primary tumors had p53 mutations. This appears to be a somewhat high incidence, even for the p53 gene. How did you exclude PCR artifacts from introducing mutations in the DNA that you analyzed?
Second, in the clinical setting, the diagnosis must often be determined before operatione from a fine-needle aspirate. Do fine-needle aspirates provide sufficient cells for DNA extraction for your analysis?
Dr. Kandioler
In response to your first question, we do PCR analysis in patients with lung cancer, and we routinely isolate DNA from the tumor and a second part of the tumor. We resect four parts from the tumor during the operation and place them in liquid nitrogen. We then isolate tumor material, normal material from the patient, and peripheral blood as negative controls to exclude artifacts.
In response to your second question, we have already done DNA sequencing analysis from bronchoalveolar lavage and biopsy material. We do not need a great amount of DNA for this analysis. If you have the advantage of first analyzing the paraffin-embedded sections from these patients' previous tumors, you know in most cases which exon carries the mutation, and you have only to analyze exactly this exon to differentiate these two tumors. The amount of DNA you will need for this analysis thus decreases further.
Footnotes
From the Departments of Surgery,a Clinical Pathology,b and Laboratory Medicine,c the University of Vienna Medical School, Vienna, Austria. 
Read at the Seventy-fifth Annual Meeting of The American Association for Thoracic Surgery, Boston, Mass., April 23-26, 1995.
References
This article has been cited by other articles:
|
|
 |
|
  Y.-W. Tung, C.-P. Hsu, S.-E. Shai, J.-Y. Hsia, S.-S. Yang, and C.-Y. Chen Surgical feasibility of ipsilateral multifocal non-small cell lung cancer in different lobes: excellent survival in node-negative subgroup Eur. J. Cardiothorac. Surg., December 1, 2003; 24(6): 1008 - 1012. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
