JTCS Speed Up Your Browser
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shackney, S. E.
Right arrow Articles by Landreneau, R. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shackney, S. E.
Right arrow Articles by Landreneau, R. J.

J Thorac Cardiovasc Surg 1999;118:259-269
© 1999 Mosby, Inc.

GENERAL THORACIC SURGERY

GENETIC EVOLUTIONARY STAGING OF EARLY NON–SMALL CELL LUNG CANCER: THE P53 -> HER-2/NEU -> ras SEQUENCE

S. E. Shackney, MD, C. A. Smith, A. Pollice, M. Levitt, MD, J. A. Magovern, MD, R. J. Wiechmann, MD, J. Silverman, MD, L. Sweeney, R. J. Landreneau, MD

From the Laboratory of Cancer Cell Biology and Genetics, Department of Human Oncology (C.A.S., A.A.P., L.S., S.E.S.) and Human Genetics (S.E.S), Department of Human Oncology (M.L.), Department of Surgery (R.J.L., J.M., R.J.W.), and Department of Laboratory Medicine (J.S.), Allegheny University of the Health Sciences, Pittsburgh, Pa.

Address for reprints: Allegheny University of the Health Sciences, Allegheny General Hospital, 320 East North Ave, Pittsburgh, PA 15212.


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Appendix: Discussion
 References
 
Introduction: The sequence of genetic evolutionary abnormalities that have occurred in a given lung cancer tumor before tumor sampling can be inferred from patterns of intracellular co-occurrence of these abnormalities in tumor cell subpopulations at the time of sampling. The same evolutionary sequences that are present within each lung cancer were evident in intertumor comparisons.
Methods: Correlated cell by cell measurements of cell DNA content, p53, Her-2/neu, and ras proteins were obtained by multiparameter flow cytometry on 46 surgically resected stage I-III primary non–small cell lung cancers. Early evolutionary changes were identified by the fact that they could appear alone in individual cells. Later appearing abnormalities were identified by the fact that they were accompanied by early abnormalities in the same cells. Patients were followed prospectively. Evolutionary patterns observed in individual tumors were correlated with subsequent clinical outcome of patients undergoing surgical resection.
Results: Three common patterns were identified: (I) a diploid DNA pathway consisting of the sequence p53 overexpression -> Her-2/neu overexpression -> ras overexpression, (II) an aneuploid DNA pathway with the same p53 -> Her-2/neu -> ras sequence, and (III) a pathway in which none of the intracellular protein measurements made here were abnormal. Fourteen tumors recurred after 11.5 months’ median study time. Nine of 12 recurrences in pathways I and II occurred in patients whose tumors were far advanced along these molecular genetic pathways.
Conclusions: Multiparameter cell-based genetic evolutionary studies may be a promising approach for identifying patients with stage I-III non–small cell lung cancer at high risk for recurrence.


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Appendix: Discussion
 References
 
Approximately 50% of patients with stage I or II non–small cell lung cancer (NSCLC) eventually have recurrent disease after surgical resection. If it were possible to identify the patients who are at high risk for recurrence at the time of surgery, they could be targeted for systemic adjuvant therapy trials, whereas those at low risk could be spared unnecessary additional treatment. Although many prognostic factors have been evaluated in NSCLC, no single prognostic factor has been able to distinguish between patients in whom relapse is likely and those in whom it is not likely with sufficient clarity to serve as the basis for reliable therapeutic decisions. On the basis of recent progress in our understanding of tumor biology, we have generated testable hypotheses regarding the initial choice and optimal use of combinations of prognostic factors for NSCLC. In keeping with a widely accepted model first proposed by Nowell,Go 1 we assume that tumors evolve from a near-normal state to aggressive malignancy by the successive overgrowth of clonal cell subpopulations that become increasingly more abnormal, owing to the progressive accumulation of multiple genetic abnormalities within individual cells. These accumulated genetic abnormalities confer various phenotypic characteristics on the cells that contain them (and on their host tumors), which include rapid growth, acquisition of the capacity to invade locally, and acquisition of the capacity to metastasize and survive at distant sites.

A recent review of patterns of DNA abnormalities and oncogene expression in human solid tumors has suggested that several sequences of genetic changes occur preferentially in individual tumors.Go 2 Common sequences of genetic evolutionary changes have recently been confirmed in human breast cancer. It was shown that ras overexpression was a relatively late finding, occurring in cells that already overexpressed Her-2/neu and/or epidermal growth factor receptor in almost all tumors studied.Go 3 The present study was prompted by the consideration that comparable patterns of genetic evolution might be observed in solid tumors of similar histologic appearance originating at different organ sites, particularly NSCLC. We would hypothesize that patients who are at high risk for recurrence are those with tumors in which abnormal cells had progressed far enough along their respective genetic evolutionary pathways to have acquired the capacity to metastasize. Conversely, patients whose tumor cells had not progressed to this extent might be at a lower risk for postoperative recurrence.

Our methodologic approach is based on the technical capability to perform multiple quantitative measurements of oncogene products on each cell in a given tumor sample by multiparameter flow cytometry. We have found that in most solid tumors, precursor populations generally persist in the background during and after the emergence of genetically more advanced tumor clones.Go Go 2-4 These clones are characterized by larger numbers of genetic abnormalities per cell and/or higher levels of expression of individual oncogene proteins.

If there is a specific order in which different genetic abnormalities appear, it will quickly become apparent when one analyzes the patterns of their co-occurrence in individual cells. Early changes will be found unaccompanied by late abnormalities in some cells. Late changes, however, will always be accompanied by the presence of early genetic changes in the same cells. A common genetic evolutionary pathway will be recognizable from the patterns of increased frequencies of occurrence of specific combinations of abnormalities in different tumors (Fig 1).



View larger version (43K):
[in this window]
[in a new window]
  		Fig. 1. Schematic illustration of two strategies for reconstructing the genetic evolutionary pathways of NSCLC. A, It is assumed that successive genetic abnormalities accumulate in individual cells and that cells representative of early and intermediate stages persist during later stages of genetic evolution. Given that a hypothetical tumor has gone through a sequence of changes consisting of A -> B -> C -> D, the intermediate cell forms that would be identified when multiple measurements are performed on each cell are A, AB, ABC, and ABCD. The sequence A -> B -> C -> D is established by the absence of cells containing B, C, or D alone, the absence of AC, AD, BC, BD, or CD in cells with only two abnormalities, and the absence of ABD, ACD or BCD in cells with three abnormalities. In this hypothetical example, the acquisition of metastatic potential by cells in the primary tumor is associated with the development of abnormality C. Patients with tumors containing cells that progressed no further than AB would be at low risk for recurrence, whereas patients with tumors containing cells with at least abnormalities ABC would be at high risk for recurrence. B, If sequence A -> B -> C -> D is recapitulated in many different tumors of the same type, then the most advanced cells in tumors that had been characterized by multiparameter analysis will exhibit only patterns A, AB, ABC, or ABCD (gray circles), and the only patients with recurrences would be those with patterns ABC or ABCD (black circles). This hypothetical example also illustrates the importance of choosing measurements that are relevant to tumor progression. If there is a subset of tumors in which the sequence of changes that drive the tumor is E -> F -> G -> H, then measurements A, B, C, and D would be uninformative.

 
We performed correlated quantitative measurements of DNA content per cell, p53 protein level per cell, and cellular levels of the Her-2/neu and ras proteins in primary tumors from 46 patients with NSCLC. In this paper we will focus on the genetic evolutionary patterns that became apparent when intertumor correlations of these genetic changes leading to oncogene product expression were analyzed.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Appendix: Discussion
 References
 
Patient population.
Tumor samples were obtained from fresh surgical specimens of primary NSCLC in patients who underwent surgical resection at Allegheny General Hospital between February 12, 1996, and April 23, 1998, with informed consent, as approved by the Allegheny General Hospital Institutional Review Board. Table I summarizes the clinicopathologic features of these patients. Median time of follow-up was 11.5 months, as of June 12, 1998. All 46 patients underwent pathologic staging as defined by the new International Lung Cancer Staging outline.Go 5


View this table:
[in this window]
[in a new window]
  		Table I. Patient characteristics
 
Sample preparation.
Freshly obtained tumor samples were finely scissor-minced in Hanks’ balanced salt solution (Gibco BRL; Life Technologies, Inc, Rockville, Md), filtered through 64 µm nylon mesh (Small Parts, Miami, Fla), and washed once with Hanks’ balanced salt solution. Cytospin slides (Shandon, Pittsburgh, Pa) were prepared from the cell suspension to confirm the presence of tumor cells in each sample, and cells were then fixed in 70% methanol for single DNA analysis and in 0.25% paraformaldehyde followed by 70% methanol for multiparameter analysis, as previously described.Go 6

Immunofluorescence staining.
Fluorescein-conjugated monoclonal antibody immunospecific for p53 protein monoclonal Clone DO-7 was purchased from Novocastra Laboratories Ltd (Newcastle upon Tyne, United Kingdom). Rabbit polyclonal antibody to c-erb -B-2, purchased from Cambridge Research Biochemicals (Cambridge, United Kingdom), was used for indirect staining. Phycoerythrin-conjugated goat anti-rabbit immunoglobulin G, purchased from Vector (Burlingame, Calif), was used as a secondary antibody. Rat monoclonal antibody to human v-H-ras , which recognizes human c-H-ras, K-ras, and N-ras, was purchased from Oncogene Science (Cambridge, Mass). This antibody was conjugated with Cy-5 (Amersham Life Sciences, Inc, Pittsburgh, Pa) and used for direct staining. JC 1939, a breast cancer cell line established in our laboratory, and for which quantitative mean levels of p53 per cell and Her-2/neu per cell were determined by enzyme-linked immunosorbent assay, was used as a quantitative staining reference for p53 and Her-2/neu immunofluorescence. Lymphocytes from healthy donors were used as low-level baseline immunofluorescence staining controls and as relative reference standards for the ras measurements.

Flow cytometry.
DAPI (4,6-diamino-2-phenylindole) (for cell DNA content measurements), FITC (fluorescein isothiocyanate), phycoerythrin, and Cy-5 fluorescence were measured with the use of an EPICS ELITE cytometer (Coulter, Miami, Fla) as previously described.Go 3

Criteria for oncogene overexpression.
Flow cytometric techniques are sufficiently sensitive to detect normal levels of p53, which are generally in the range of 5000 to 7500 molecules per cell. In the present study, intracellular levels of p53 exceeding 10,000 molecules per cell were considered abnormal, and tumors in which the mean level of p53 overexpression per cell exceeded 10,000 molecules per cell were classified as p53 overexpressors. For Her-2/neu, tumors exhibiting mean levels exceeding 150,000 molecules per cell were classified as overexpressors, and tumors exhibiting mean levels exceeding 300,000 molecules per cell were classified as high overexpressors. Tumors exhibiting mean levels of ras per cell that were more than twice the lymphocyte reference (mean > 20,000 units per cell) were classified as overexpressors, and tumors exhibiting mean levels of ras per cell that were more than 4 times the lymphocyte reference (mean > 40,000 units per cell) were classified as high overexpressors.

Statistical analysis.
Means of 2 groups were compared with the use of the Student t test. Associations between variables were assessed by the {chi}2 test or by Fisher’s exact test when appropriate. Kaplan-Meier survival curves were compared by log rank analysis.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Appendix: Discussion
 References
 
Abnormal mean p53 expression and ploidy.
Elevated overall mean p53 levels per cell were observed in 35 (76%) of 46 cases of NSCLC. This may be an underestimate inasmuch as p53 overexpression is observed in the presence of only about 85% to 90% of p53 mutations. Stop codon mutations and deletions result.Go Go 7,8

Among the 11 tumors that did not exhibit p53 overexpression, 8 were diploid by flow cytometry. Five of the diploid tumors without elevated p53 levels were distinguished by the absence of overexpression of any of the oncogene proteins evaluated in the present study, suggesting that these 5 tumors followed a genetic evolutionary pathway separate from that of all the other tumors. The absence of p53 overexpression was uncommon in aneuploid tumors, occurring in only 3 of 21 cases. Two of these aneuploid tumors showed no other oncogene protein abnormalities as well, suggesting that they might also be following a separate genetic evolutionary pathway.

Abnormal mean Her-2/neu expression and ploidy.
Mean levels of Her-2/neu per cell that exceeded 150,000 molecules per cell were observed in the diploid and/or aneuploid components of 22 (48%) of 46 tumors. The highest mean levels of Her-2/neu levels among the aneuploid components of aneuploid tumors by several-fold. However, Her-2/neu was rarely overexpressed in the diploid components of aneuploid tumors. These two findings suggest that the molecular mechanisms that underlie Her-2/neu overexpression may be different in diploid and in aneuploid tumors. Furthermore, the fact that the diploid components of aneuploid tumors frequently overexpressed p53 but not Her-2/neu suggests that p53 overexpression preceded Her-2/neu overexpression in many of these tumors.

Abnormal mean ras expression and ploidy.
Elevated mean levels of ras per cell were observed in the diploid or aneuploid components of 25 (54%) of 46 tumors. High mean levels of ras per cell were observed both in diploid tumors and in the aneuploid components of aneuploid tumors. Like Her-2/neu, ras was rarely overexpressed in the diploid components of aneuploid tumors, suggesting that this abnormality often developed later than p53 overexpression in aneuploid tumors.

Abnormal mean p53 expression and mean Her-2/neu overexpression.
There was a relationship between mean p53 level per cell and mean Her-2/neu level per cell in the same tumor. Among 35 tumors with elevated mean p53 levels per cell, 20 tumors also had elevated mean Her-2/neu levels per cell (>150,000 molecules per cell); among 11 tumors in which mean p53 levels were not elevated, only 2 had mean Her-2/neu levels per cell that exceeded 150,000 molecules per cell. The difference was statistically significant (P = .038 by Fisher’s exact test).

Abnormal mean p53 expression and mean ras overexpression.
There also appeared to be a relationship between mean p53 level per cell and mean ras level per cell in the same tumor. Among 35 tumors with elevated mean p53 levels per cell, 25 tumors also had elevated mean ras levels per cell (>20,000 molecules per cell); among 11 tumors in which mean p53 levels were not elevated, only 4 had mean ras levels per cell that exceeded 20,000 molecules per cell. However, this difference did not achieve statistical significance (P = .07 by Fisher’s exact test).

Abnormal mean Her-2/neu expression and mean ras overexpression.
There was a strong association between mean Her-2/neu level per cell and mean ras level per cell in the same tumor. Among 22 tumors with mean Her-2/neu levels exceeding 150,000 molecules per cell, 19 tumors also had elevated mean ras levels per cell; among 24 tumors in which mean Her-2/neu levels were not elevated, only 10 had mean ras levels per cell that exceeded 20,000 molecules per cell. The difference was statistically significant (P = .002 by Fisher’s exact test).

Intratumor cell by cell correlations.
When the patterns of oncogene expression were examined within individual cells in each tumor, the most common pattern observed was one in which cells that overexpressed p53 only were present in abundance, Her-2/neu overexpression was commonly restricted to a subset of cells that also overexpressed p53, and ras overexpression was observed almost exclusively in cells that also overexpressed both Her-2/neu and p53 proteins. All of these relationships are illustrated in the example shown in Fig 2. In addition, in tumors that contained an aneuploid component, the cells that overexpressed all three proteins were predominantly aneuploid. In a subset of predominantly diploid tumors that did not exhibit p53 overexpression, the majority of cells often did not express Her-2/neu or ras either.



View larger version (81K):
[in this window]
[in a new window]
  		Fig. 2. Multiparameter analysis of an NSCLC, illustrating the relationships among p53, Her-2/neu, ras, and aneuploidy in the same cells. Reference lines indicate diploid cell DNA content (vertical line) and 10,000 molecules per cell, the upper limit for normal mean level of p53 per cell (horizontal line). A, A bivariate distribution of cell DNA content and p53 protein content per cell. This sample contains a diploid cell population with normal levels of p53 per cell (arrow a), a hyperdiploid cell population (arrow b) with slightly increased levels of p53 protein per cell, and a heterogeneous aneuploid cell population with substantial overexpression of p53 protein per cell (bracket c). B, A bivariate distribution of p53 protein content per cell and Her-2/neu protein content per cell. The horizontal reference line for Her-2/neu is at 150,000 molecules per cell and the vertical reference line for p53 is at 10,000 molecules per cell. This sample contains a large cell subpopulation that overexpresses p53 alone (arrow d) but does not contain cells that overexpress Her-2/neu in the absence of p53 protein (arrow e), indicating that p53 overexpression preceded Her-2/neu in the evolutionary sequence in this tumor. All of the cells that express ras at high levels per cell (dots) are part of the subset of cells that also overexpress p53 and Her-2/neu (arrow f), indicating that ras is last in the sequence.

 
Genetic pathways.
The foregoing relationships suggest a common sequence of genetic evolutionary changes in which p53 abnormalities develop first, followed by Her-2/neu overexpression in a subset of p53 overexpressing cells. The overexpression of ras would appear to occur last, commonly in a subset of cells that had developed p53 and Her-2/neu abnormalities previously.

In Fig 3, individual tumors, represented by shaded circles, are mapped to the pathway that most closely reflects their individual patterns of oncogene expression, and each tumor is placed at a point that corresponds most closely to the degree of its advancement along that pathway. The patterns traced out by these tumors appear to define three major genetic evolutionary pathways in NSCLC. The most prominent of these (area II) is an aneuploid pathway that features p53 overexpression, and marked Her-2/neu and ras overexpression in its later stages. The second is a diploid pathway (area I) that also features p53 overexpression and marked Her-2/neu and ras overexpression in its later stages. This pathway includes a minor diploid pathway in which p53 overexpression is not observed, but in which Her-2/neu and ras are overexpressed. The third distinct pathway is a diploid pathway (area III) in which mean p53, Her-2/neu, and ras levels are all normal, as previously noted.



View larger version (54K):
[in this window]
[in a new window]
  		Fig. 3. A reconstruction of the genetic evolutionary pathways in NSCLC based on intertumor comparisons. Gray circles represent individual tumors; black circles indicate tumors that recurred. Three pathways are identified. Recurrences are found among the most advanced cells in pathways I and II. Pathway III was not subject to sequence analysis, because none of the measurements performed in this study were relevant to this pathway.

 
There were 14 tumor recurrences (30.4%) after a median time on study of only 11.5 months. These are shown in Fig 3Go as black circles. It is apparent that recurrences are observed along all three of the major pathways and that, at least with respect to the first two pathways, all but 1 of the patients who had early recurrence had primary tumors that were relatively far advanced along their respective genetic evolutionary pathways, as evidenced by expression of multiple oncogene products. Two recurrent tumors classified as non-overexpressors had mean Her-2/neu levels that were between 100,000 and 150,000 molecules per cell. Thus the threshold accepted for elevation in the expression of this oncogene may have underestimated its importance for these tumors. It is also of interest that the majority of adenocarcinomas and undifferentiated large cell carcinomas were aneuploid tumors (16/29, or 55%), whereas the majority of squamous cell carcinomas were diploid tumors (9/17, or 53%). However, this association was found not to be statistically significant (P = .09).

Tumor recurrence in relation to clinicopathologic and/or genetic stage of tumor advancement was assessed. Tumor recurrences were observed among patients at all stages of disease and in comparable proportions of patients at each stage (Table II). There were no statistically significant differences in disease-free survival by stage (Fig 4). However, the number of cases in this study was relatively small, and the duration of follow-up to date was relatively short. Most of the recurrent tumors participating in pathways I and II were in advanced stages of their genetic evolution (Fig 3Go). Among the 12 recurrences in pathways I and II, 11 overexpressed ras and 9 overexpressed both Her-2/neu and ras. If the tumors with mean Her-2/neu levels in the range of 100,000 to 150,000 molecules were to be included among the Her-2/neu overexpressors, then 11 of the 12 recurrences in pathways I and II would be classified as overexpressors of both Her-2/neu and ras. The twelfth tumor exhibited a borderline increase in mean p53 level per cell and normal mean Her-2/neu and ras levels; it is conceivable that this tumor actually may have been following pathway III. There was no obvious correlation between clinicopathologic stage of disease and the degree of tumor genetic evolutionary progression (Table III). Within each genetic evolutionary pathway, genetically advanced stage I tumors were found together with genetically advanced stage III tumors; similarly, stages I, II, and III were all represented among subsets of tumors that had not advanced very far in the genetic evolutionary sequence. The data in Table IIIGo raise the possibility that recurrences among patients with stage I disease may occur more commonly in tumors that follow pathway I, and that they may occur later. It will be necessary to study a larger number of cases for a longer period of time to determine the validity of this early observation.


View this table:
[in this window]
[in a new window]
  		Table II. Tumor recurrences in relation to stage of disease
 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 4. Disease-free survival by clinicopathologic stage (stages I-III). There were no statistically significant stage-related differences in disease-free survival.

 

View this table:
[in this window]
[in a new window]
  		Table III. Tumor recurrences in relation to genetic evolutionary pathway, genetic evolutionary stage, and clinicopathologic stage of disease
 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Appendix: Discussion
 References
 
In this article we have presented evidence suggesting that p53 protein overexpression, Her-2/neu overexpression, and ras overexpression often occurred together in the same tumor cells in various combinations in the majority of cases of NSCLC. Among the three protein abnormalities studied, p53 overexpression was the only abnormality that was commonly observed in the absence of the other two in individual tumors and, more specifically, within the individual diploid cell components of individual tumors. This suggested that p53 abnormalities may have preceded the others in the genetic evolutionary sequence that leads to aggressive malignancy. This is consonant with published studies that indicate that p53 mutations (which are usually associated with p53 overexpression) often occur in preinvasive stages of NSCLC.Go Go 9-11 Ras gene mutations have been reported to occur at very early preneoplastic stages of development in NSCLC in some studies,Go 12 but not in others.Go 13 However, overexpression of the ras protein has been found to be a late phenomenon.Go 14 Studies in experimental tumor systems suggest that the introduction of functional wild type p53 can suppress the transforming effects of ras overexpression in the same cells,Go 15 suggesting the possibility that in spontaneous human tumors it is only when wild type p53 is neutralized that ras overexpressing cells can flourish. This premise is supported by the results of the present study. Wild type p53 inactivation can occur within individual cells in a number of ways and might not be accompanied by overexpression of the p53 protein. Although most missense p53 mutations result in a form of the p53 protein that accumulates in cells, gene mutations that result in truncation of the p53 protein can lead to its underexpression or its total absence. Loss of wild type p53 function can occur in the absence of p53 mutation or deletion, as a secondary or tertiary effect of other abnormalities such as those involving MDM2Go Go 16,17 or p19ARF.Go 18 Thus it is likely that some p53 abnormalities went undetected in the present study, particularly in tumors in which stigmata of wild type p53 inactivation were present, but p53 protein levels were not elevated.

Loss of wild type p53 function leads to a distinctive form of genetic instability, which is often manifested by the development of gross aneuploidy and multiple chromosome breaks that often result in deletions, translocations, and gene amplification.Go 2

An association was observed in the present study between p53 overexpression and Her-2/neu overexpression in individual tumors. Her-2/neu gene amplification is relatively common in some human tumors, such as breast cancer,Go Go 19,20 and is often associated with levels of Her-2/neu overexpression that are in excess of 500,000 molecules per cell.Go Go 21,22 However, Her-2/neu amplification has been reported to be a relatively uncommon mechanism for Her-2/neu overexpression in NSCLC.Go 23 The molecular basis for Her-2/neu overexpression in the cases reported here remains to be elucidated.

A strong relationship between Her-2/neu overexpression and ras overexpression was observed in the present study. Within individual cells in individual tumors, ras overexpression was rarely observed in cells that did not also overexpress Her-2/neu and/or p53 protein overexpression. This suggests that ras overexpression is a relatively late evolutionary change. Similar findings have also been reported in human breast cancer.Go Go 3,24

On the basis of the patterns of overexpression of p53, Her-2/neu, and ras in individual tumors, three common genetic evolutionary pathways for non–small cell could be identified. The most common was an aneuploid pathway in which the p53 -> Her-2/neu -> ras sequence predominated. A similar pathway was found in diploid tumors. A third pathway was identified in which none of the proteins measured here was overexpressed. Early recurrences were observed in all three pathways, and at least with respect to the first two pathways, recurrences were found largely among tumors that had reached relatively advanced stages of this genetic evolutionary schema. These genetic patterns appear to be prognostic variables independent from clinicopathologic stage of NSCLC.

The findings of this investigation appear promising in identifying patients at risk for early recurrence despite favorable clinicopathologic staging. These molecular genetic changes may also identify patients with more advanced clinicopathologic staging who have a more favorable disease-free survival. These data must be interpreted with caution, in view of the small number of patients studied and the relatively brief duration of follow-up. However, what may be more important than these interim results is the method by which they were obtained. This approach uses current understanding of the molecular mechanisms for tumor evolution to develop biologic risk models for recurrent disease. This approach can provide direction for future molecular biologic studies to further define combinations of genetic prognostic factors associated with stage independent virulence of NSCLC.


   Appendix: Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Appendix: Discussion
References
 
Dr Eric Vallieres (Seattle, Wash). Thank you for suggesting an interesting concept on the genetic evolutionary sequence in NSCLC.

The field of molecular biology and genetics has grown substantially in the past decade, owing mainly to the discovery of various genetic abnormalities, but also to the development of different and various reproducible molecular biology techniques that allow us to identify and measure these proteins.

The potential applications of these discoveries are innumerable and could affect the ways we prevent, diagnose, prognosticate, or treat some of the tumors that we see.

In this retrospective study of 46 stage I to III resected NSCLCs, you have measured the cellular contents of DNA, p53 -> Her-2/neu -> K-ras proteins. From your observations of the various presences of zero, one, two, or all three of the proteins, you have postulated the following evolutionary genetic sequence: p53 -> Her-2/neu -> K-ras. Secondarily, you describe a possible additional prognostic tool that could be used in treating these tumors. Please allow me a few comments and questions.

Before reviewing your abstract, I had never heard of the notion of measuring the cellular contents of these proteins. To me, this appears to be a new technology and a new tool. Can you compare it with other means of identifying these various proteins? How does your technique compare with these other techniques? How do we know that what you are measuring are surrogates for mutations of these proteins?

Dr Landreneau. This is a prospective study that uses an existing technology in new ways. Flow cytometry has been used to quantitate cellular DNA content using stoichiometric DNA-binding fluorescent dyes and to quantitate cell surface proteins using immunofluorescent antibodies since the late 1970s. The new uses for flow cytometry in this study are as follows: (1) It combines cell fixation techniques that cross-link intracellular proteins (to prevent their leakage out of cells) and then make the cell membrane permeable to perform valid quantitative intracellular immunofluorescent measurements on each cell; (2) it focuses on the measurement of oncogene protein products that accumulate in individual cells during the course of tumor progression; (3) it performs multiple measurements on each cell; and (4) it quantitates as many of these protein products as possible in absolute units (molecules per cell) using appropriate reference cells.

Dr Vallieres. How does your technique compare with these other techniques?

Dr Landreneau. Each technique has advantages and disadvantages. In our view, the main advantages of multiparameter flow cytometry (MP-FCM) in comparison with DNA- or RNA-based techniques are as follows: (1) MP-FCM measurements are performed one cell at a time on large numbers of cells, preserving intracellular relationships among the measured quantities and avoiding the signal averaging effects that are inherent in methods that require the pooling of material from large numbers of disrupted cells, and (2) they focus on protein abnormalities that are likely to bear directly on pathogenesis, rather than on genetic mechanisms, which may be more relevant to etiology. In comparison with other protein-based measurements, and specifically immunohistochemistry, MP-FCM appears to be more sensitive, at least with respect to the detection of p53 protein and H-2/neu protein. Normal p53 protein levels (~5000 molecules per cell) are usually undetectable by immunohistochemistry, whereas MP-FCM can detect as little as 1000 to 2000 molecules per cell. Her-2/neu protein levels are usually detected by immunohistochemistry when they exceed 200,000 to 500,000 molecules per cell (normal levels, 20,000-80,000 molecules per cell). MP-FCM can detect Her-2/neu reliably at levels in the range of 50,000 molecules per cell or more.

Dr Vallieres. How do we know that what you are measuring are surrogates for mutations of these proteins?

Dr Landreneau. The relationship between cell protein content and gene mutation is not a simple one. In the case of p53, the key issue is abrogation of wild type p53 function, which can occur by a variety of different mechanisms. Mutations of the p53 gene are common, and approximately 85% of mutations are missense mutations with clear-cut p53 protein overexpression. However, abrogation of wild type p53 function can occur by mechanisms that involve neither p53 mutation nor p53 protein overexpression (eg, MDM2 or p19ARF gene abnormalities, or p53 allelic loss). As a result, no matter which single measurement one performs, one is likely to miss some p53 abnormalities that result in loss of p53 function. With regard to Her-2/neu abnormalities, gene amplification is common in breast cancer, but not in lung cancer, so that Her-2/neu overexpression is likely to be due to indirect effects of abnormalities in other genes that might regulate Her-2/neu levels.

Dr Vallieres. Others have previously correlated the presence of these genetic markers with a history of tobacco exposure. We know, for example, that K-ras point mutations have been measured in a third of adenocarcinomas and in individuals who have smoked or who are smoking, but fewer than 5% of adenocarcinomas are seen in nonsmokers, and similar types of observations have been done with p53.

Do you have data concerning tobacco exposure on these 46 patients and tumors?

Dr Landreneau. All but 2 of the patients were heavy smokers or former smokers (mean 54 pack-years, range 20-150 pack years).

Dr Vallieres. In previous studies also, K-ras mutations were found in up to 30% of adenocarcinomas but were not seen in the other subtypes of NSCLC. On the other side, p53 is identified in 40% of adenocarcinomas but 70% of squamous cell carcinomas. Have you analyzed the profile of these 46 tumors according to their subtypes of NSCLC?

Dr Landreneau. MP-FCM measured intracellular ras protein content, not gene mutations, and the antibody used reacted with H-ras and N-ras protein as well (although most of the ras detected in this study was likely to be K-ras ). The distinction between ras protein overexpression and ras mutations is an important one, inasmuch as there is not a 1:1 correspondence between the two. In breast cancer, for example, ras protein overexpression is common, but ras gene mutations are not. Furthermore, K-ras mutations can occur at such early premalignant stages of the development of solid tumors that their prognostic usefulness might be open to some question. In the present study, ras protein overexpression was found to occur as a late phenomenon (commonly after the development of p53 and Her-2/neu abnormalities). It is conceivable that ras protein overexpression might reflect abnormalities in the intracellular functional state of ras more closely than the presence or absence of mutation of the gene. In this regard, a number of published studies in experimental tumor systems have shown that ras overexpression is more closely associated with the acquisition of metastatic potential than the presence or absence of ras mutations.

Dr Vallieres. Were K-ras mutations seen in non-adenocarcinoma tumors?

Dr Landreneau. With regard to patterns of p53 and ras expression by tumor subtype in the present study, the frequencies of p53 overexpression were similar in adenocarcinomas and squamous cell carcinomas (17/19 or 89% of adenocarcinomas vs 14/18 or 78% of squamous cell cancers) and a little lower in large cell undifferentiated tumors (6/9 or 67%). The frequencies of ras overexpression were also similar in adenocarcinomas and squamous cell carcinomas (14/19 or 74% of adenocarcinomas vs 11/18 or 61% of squamous cell cancers) and somewhat lower in large cell undifferentiated tumors (4/9 or 44%). We should not try to read too much into these frequency data, because the number of patients is still relatively small.

Dr Vallieres. The next comment is that we know that in small cell lung cancers p53 abnormalities will be seen in 80% of these tumors, but K-ras and Her-2/neu abnormalities are infrequent. Can you comment on that?

Dr Landreneau. Although we did not examine small cell lung tumors in this study, our findings in NSCLC appear to be different from those reported in the literature for small cell lung cancer, in which Her-2/neu and ras abnormalities are uncommon. We suspect that this may be due to the loss of Rb protein in small cell lung cancer. The mitogenic signaling pathway ordinarily proceeds from the activation of membrane signaling receptors like Her-2/neu though activation of ras , raf , MEK, and MAPK, the induction of cyclin D1, cyclin D/cdk4-mediated phosphorylation of Rb, and the release of free E2F1, which leads to a round of DNA synthesis and cell division. In tumors in which Rb protein is intact (which includes the majority of NSCLCs), a strategy of chronic mitogenic stimulation might account for many known observations (such as epidermal growth factor receptor or Her-2/neu overexpression, ras overexpression, cyclin D1 overexpression, and/or loss of the cyclin D1/cdk4 inhibitor, p16). However, one might expect that when Rb protein is lost or inactivated (as in nearly all small cell lung cancers), the linkage between early steps in the mitogenic signaling pathway and E2F1-mediated DNA synthesis would be disrupted, rendering any abnormalities in these early steps irrelevant to cell cycle progression. Thus we would speculate that in Rb-negative small cell lung cancer, Her-2/neu overexpression and/or ras abnormalities would offer no proliferative advantage, and they would not be favored by evolutionary selection.

Dr Vallieres. What percentage of the tumors had both K-ras and p53?

Dr Landreneau. On the basis of the criteria that we used for distinguishing high levels of expression of p53 and ras from low levels of expression, 25 (54%) of 46 tumors contained cells that overexpressed both in the same cells.

Dr Vallieres. This probably reflects a difference in the technique of measuring, because previously both proteins were measured in only about 10% of these tumors.

This evolutionary concept is new and interesting. Whether the sequence is real or whether the simultaneous presence of these three or even more abnormalities only reflects a direct effect of carcinogen exposure without sequencing needs to be clarified. The identification of these three proteins and resected tumors appears to identify bad plain tumors and could potentially serve as an additional prognostic tool to help guide therapy. Identification of these proteins on pre-resected sampling could potentially lead to interesting induction strategies and therapies. However, as you stated yourself, this study is early, it is small, and the follow-up remains relatively short.


   Footnotes
 
Read at the Twenty-fourth Annual Meeting of The Western Thoracic Surgical Association, Whistler, British Columbia, June 24-27, 1998. Back


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Appendix: Discussion
 References
 

  1. Nowell P. The clonal evolution of tumor cell populations. Science 1976;194:23-8. [Abstract/Free Full Text]
  2. Shackney S, Shankey T. Common patterns of genetic evolution in human solid tumors. Cytometry 1997;29:1-27. [Medline]
  3. Shackney S, Pollice A, Smith C, Janocko L, Sweeney L, Brown K, et al. Intracellular coexpression of epidermal growth factor receptor, Her-2/neu, and p21ras in human breast cancers: evidence for the existence of distinctive patterns of genetic evolution that are common to tumors from different patients. Clin Cancer Res 1998;4:913-8. [Abstract]
  4. Shackney S, Smith C, Pollice A, Janocko L, Singh S, Groft D, et al. Preferred genetic evolutionary sequences in human breast cancer: a case study. Cytometry 1995;21:6-13. [Medline]
  5. Mountain CF. Revisions in the International System for Staging Lung Cancer. Chest 1997;111:1710-7. [Abstract/Free Full Text]
  6. Pollice AA, McCoy JP Jr, Shackney SE, Smith CA, Agarwal J, Burholt DR, et al. Sequential paraformaldehyde and methanol fixation for simultaneous flow cytometric analysis of DNA, cell surface proteins, and intracellular proteins. Cytometry 1992;13:432-44. [Medline]
  7. Bodner S, Minna J, Jensen S, D’Amico D, Carbone D, Mitsudomi T, et al. Expression of mutant p53 proteins in lung cancer correlates with the class of gene mutation. Oncogene 1992;7:743-9. [Medline]
  8. Levine A. p53, the cellular gatekeeper for growth and division. Cell 1997;88:323-31. [Medline]
  9. Sundaresan V, Ganly P, Hasleton P, Rudd R, Sinha G, Bleehen N, et al. p53 and chromosome 3 abnormalities, characteristic of malignant lung tumors, are detectable in preinvasive lesions of the bronchus. Oncogene 1992;7:1989-97. [Medline]
  10. Sozzi G, Miozzo M, Donghi R, Pilotti S, Caraini C, Pastorino U, et al. Deletions of 17p and p53 mutations in preneoplastic lesions of the lung. Cancer Res 1992;52:6079-82. [Abstract/Free Full Text]
  11. Bennett W, Colby T, Travis W, Borkowski A, Jones R, Lane D, et al. p53 protein accumulates frequently in early bronchial neoplasia. Cancer Res 1993;53:4817-22. [Abstract/Free Full Text]
  12. Li Z-H, Zheng J, Weiss L, Shibata D. c-K-ras and p53 mutations occur very early in adenocarcinoma of the lung. Am J Pathol 1994;144:303-9. [Abstract]
  13. Sugio K, Kishimoto Y, Virmani A, Hung J, Gazdar A. K-ras mutations are a relatively late event in the pathogenesis of lung carcinomas. Cancer Res 1994;54:5811-5. [Abstract/Free Full Text]
  14. Dosaka H, Harada M, Kuzumaki N, Kobayashi H, Miyamoto H, Kawakami Y. The relationship of clinical classification to ras p21 expression in human non–small cell lung cancer. Oncology 1988;45:396-400. [Medline]
  15. Finlay C, Hinds P, Levine A. The p53 proto-oncogene can act as a suppressor of transformation. Cell 1989;57:1083-93. [Medline]
  16. Haupt Y, Maya R, Kazaz A, Oren M. Mdm2 promotes the rapid degradation of p53. Nature 1997;387:296-99. [Medline]
  17. Kubbutat M, Jones S, Vousden K. Regulation of p53 stability by Mdm2. Nature 1997;387:299-303. [Medline]
  18. Pomerantz J, Schreiber-Agus N, Liegeois N, Silverman A, Alland L, Chin L, et al. The Ink4a tumor suppressor gene product, p19Arf, interacts with MDM2 and neutralizes MDM2’s inhibition of p53. Cell 1998;92:713-23. [Medline]
  19. Iglehart J, Kraus M, Langton B, Huper G, Kerns B, Marks J. Increased ErbB-2 gene copies and expression in multiple stages of breast cancer. Cancer Res 1990;50:6701-7. [Abstract/Free Full Text]
  20. Ciocca D, Fujimura F, Tandon A, Clark G, Mark C, Lee-Chen G-J, et al. Correlation of Her-2/neu amplification with expression and with other prognostic factors in 1103 breast cancers. J Natl Cancer Inst 1992;84:1279-81. [Free Full Text]
  21. Press M, Pike M, Chazin V, Hung G, JA U, Markowicz M, et al. Her-2/neu expression in node-negative breast cancer: direct tissue quantitation by computerized image analysis and association of overexpression with increased risk of recurrent disease. Cancer Res 1993;53:4960-70. [Abstract/Free Full Text]
  22. Robertson K, Reeves J, Smith G, Keith W, Ozanne B, Cooke T, et al. Quantitative estimation of epidermal growth factor receptor and c-ErbB-2 in human breast cancer. Cancer Res 1996;56:3823-30. [Abstract/Free Full Text]
  23. Schneider P, Hung M-C, Chiocca S, Manning J, Zhao X, Fang K, et al. Differential expression of the c-ErbB-2 gene in human small cell and non–small cell lung cancer. Cancer Res 1989;49:4968-71. [Abstract/Free Full Text]
  24. Shackney S, Pollice A, Smith C, Alston L, Singh S, Janocko L, et al. The accumulation of multiple genetic abnormalities in individual tumor cells in human breast cancers: clinical prognostic implications. Cancer J 1996;2:105-14.
Received for publication July 15, 1998. Revisions requested Oct 12, 1998; revisions received Feb 9, 1999. Accepted for publication March 30, 1999.


This article has been cited by other articles:


Home page
 Arch SurgHome page
R. Y. Chandawarkar and D. P. Guyton
Oncologic Mathematics: Evolution of a New Specialty
Arch Surg, December 1, 2002; 137(12): 1428 - 1434.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
I. Osman, H. I. Scher, M. Drobnjak, D. Verbel, M. Morris, D. Agus, J. S. Ross, and C. Cordon-Cardo
HER-2/neu (p185neu) Protein Expression in the Natural or Treated History of Prostate Cancer
Clin. Cancer Res., September 1, 2001; 7(9): 2643 - 2647.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to Personal Folders
Right arrow Download to citation manager
Right arrow Permission Requests
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Shackney, S. E.
Right arrow Articles by Landreneau, R. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Shackney, S. E.
Right arrow Articles by Landreneau, R. J.

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