| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2000;119:1119-1125
© 2000 The American Association for Thoracic Surgery
GENERAL THORACIC SURGERY |
From the Department of Surgery II,a Clinical Analysis Department (Tumoral Biology)b and Epidemiology Unit,c Clínico Universitario San Carlos Hospital, Complutense University of Madrid, Madrid, Spain.
Address for reprints: Antonio J. Torres García, MD, Departamento de Cirugía II, Hospital Clínico Universitario San Carlos, C/Martín Lagos S/N, 28040 Madrid, Spain (E-mail: AJTORRES{at}teleline.es ).
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methods Results Discussion References |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methods Results Discussion References |
|---|
In the past few years, research has focused on the search for new biologic factors of prognostic value.
The oncogene neu was identified by transfection of NIH3T3 cells with DNA from rat neuroblastomas and glioblastomas induced by perinatal treatment with the alkylating agent nitrous-ethyl-urea. These tumors are associated with expression of a 185-kd protein on the surface of the tumoral cells.
1
To date, the p185neu protein has been determined by use of immunohistochemistry or Western blot analysis. These methods give semiquantitative results and are associated with a certain degree of subjectivity. Quantification of the p185neu protein by means of enzyme-linked immunosorbent assay (ELISA) has recently been accepted. Theoretically, this procedure for quantification of p185neu offers validated characteristics 2 that make it a good candidate for use in clinical practice.
In the light of these observations, this study was designed to assess the relationship between p185neu protein and overall survival in NSCLC.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methods Results Discussion References |
|---|
3 in the Clinic Hospital (10 women and 92 men; mean age, 62 ± 9 years; age range, 31-83 years) from October 1990 to June 1994. The surgical team was the same in all patients. Seventy lobectomies, 3 bilobectomies, and 29 pneumonectomies were performed; mediastinal node dissection was performed in all patients. The distribution of the 102 patients in relation to their histologic type in accordance with the World Health Organization classification 4 was as follows: 64 (63%) patients were classified as having epidermoid carcinoma, 30 (30%) adenocarcinoma, and 8 (7%) undifferentiated large cell tumors. Pathologic staging according to the TNM classification 3 was as follows: 58 (57%) patients in stages IA to IB, 8 (8%) in stages IIA to IIB, and 36 (35%) in stage IIIA. On analysis of the degree of tumoral differentiation, the tumors were allocated to one of three groups 5: 37 (32%) poorly differentiated tumors, 55 (48%) moderately differentiated tumors, and 23 (20%) well-differentiated tumors. Before surgical treatment, none of the patients received chemotherapy or radiotherapy. To assess a possible prognostic value of overexpression of p185, we performed postoperative follow-up studies 3 times monthly for the first 2 years and every 6 months in the following years. Mean follow-up of this group was 26 months (SD, 18 months; range, 1-70 months). Of all the patients in this group, 7 (7%) died in the immediate postoperative period. During follow-up, 3 patients died of unrelated causes, and 40 (40%) patients died of the lung cancer.
At the time of this study, tumoral relapse had been detected in 41 (40%) of the 102 patients who underwent tumor resection for curative purposes.
Twenty patients who had received surgical treatment for idiopathic pneumothorax were included to establish control levels of p185 expression as a reference value. This group comprised 5 women and 15 men with a mean age of 48 years. The 3 specimen types were compared through a cumulated distribution of p185neu protein (Fig 1).
|
For p185neu protein determinations, samples were processed as followed. First they were weighed and ground, and then the samples were homogenized and centrifuged to obtain a supernatant. This supernatant was ultracentrifuged to obtain the cytosol, which was distributed into aliquots for p185neu protein determination. All the assays were blinded to the clinical results.
p185neu protein assay
Determination of the p185 protein was performed by use of ELISA with reagents supplied by Oncogene Science (Uniondale, NY). ELISA consists of a sandwich technique requiring two antibodies: one mouse monoclonal antibody to capture the antigen and a polyclonal rabbit antibody for its detection. The latter recognizes the external domain of the p185 protein. 6
The total cytosol protein concentration was assayed by the method of Lowry, Rosenbrought, and Farr. 7 Samples were diluted before p185 determination to ensure a total protein concentration of 2 to 4 mg/mL; p185 results were expressed as units per milligram of protein.
Statistical analysis
This was a prospective cohort study, with survival as the dependent variable. The protocol of independent variables studied individually was tumoral stage, histologic type, degree of differentiation, atypical index, ploidy, age, and oncoprotein p185 determinations.
Survival was estimated according to the method of Kaplan and Meier 8 for independent variables to study the outcome death. The cut-off value was established from the distribution percentiles; the 50th percentile or between the 50th and 80th percentile values was used to categorize patients. The survival curves of each subgroup were compared by the Breslow exact test.
Finally, the Cox proportional risk regression model was fitted to data to estimate the independent prognostic importance of p185 protein. The included variables were those variables analyzed in univariable study (P < .15). All interactions were tested according to a hierarchic model. Adjusted hazard ratios are presented with their corresponding confidence intervals (CIs) at 95%. The model’s basic assumptions were evaluated (proportional hazards). In each hypothesis contrast a type I (
) error less than .05 of the null hypothesis was rejected.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methods Results Discussion References |
|---|
|
).
Mean values of the p185 protein in these 3 groups showed a clear difference between the control group and the group with a tumor (healthy tissue, P = .01, and tumoral tissue, P = .004; Fig 1).
The p185 concentration and the distribution of the results according to the patients’ characteristics are given in Table II. Statistically significant differences were not found in the p185 distribution in relation to histologic type, tumoral stage, the degree of tumoral differentiation, atypia index, or patient age. However, aneuploid samples were most prevalent and had significantly higher p185 levels than euploid samples (P = .01).
|
To study survival in relation to p185 values, the p185neu 50th percentile (210 U/mg) and 80th percentile (350 U/mg) values were used. Because significant differences were not found between the less than 210 U/mg range and the 210 U/mg to 349 U/mg range, the 80th percentile was taken as the sole reference value (Fig 2).
|
350 U/mg) showed patients with p185neu of 350 U/mg or greater to have a worse prognosis (P = .03), with a median survival of 16 months. At 24 months, 60% of patients with p185neu values of less than 350 U/mg survived. Significant differences in stratified analysis were found between levels of p185 overexpression in the epidermoid-type cancers, with survival being significantly different in groups of patients with p185neu values of 350 U/mg or greater and less than 350 U/mg (P = .03; Fig 3).
|
When tumoral stages were studied individually, global survival significantly declined in stage I patients with values of p185neu of 350 U/mg or greater (P = .0003; Fig 4). The median survival of stage I patients with values of p185neu of 350 U/mg or greater was 25 months (95% CI, 1-49 months). At this time, 84% of stage I patients with a p185neu value of less than 350 U/mg were alive.
|
In the study, according to the degree of tumoral differentiation, survival significantly declined in patients with poorly differentiated tumors with a p185 expression above the 80th percentile (P = .05).
When studying the effect of p185 expression on survival in patients with euploid and aneuploid tumors (flow cytometric DNA analysis 9), survival was significantly lower in euploid tumors with values of p185neu of 350 U/mg or greater (P = .04). No significant differences were found in survival in relation to p185 expression in patients with aneuploid tumors (P = .22).
In the study on the behavior of the population according to the degree of atypia, there was no significant association between survival and p185 values. However, p185 levels were significantly different (P = .01) in patients over 65 years old. This group had a median survival of 9 months (95% CI, 4-14 months). In patients younger than 65 years, p185 values had no statistically significant influence on survival (P = .22).
The p185neu variable was studied as (1) categorized and (2) quantified (Table III):
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and methods Results Discussion References |
|---|
10
In a previous study the authors carried out a validation study of the ELISA technique, in which p185 determination was done by using ELISA and simultaneously by using immunohistochemistry. The two techniques were found to be well correlated (r = 0.85, P < .01). 11
Recent studies have also demonstrated a clear correlation between p185 determination by means of enzyme immunoassay and immunohistochemistry. 12,13 Both techniques can be used complementarily to measure p185 values in tissue. Determination by ELISA rectifies the subjectivity of the immunohistochemical technique and provides a more precise distinction between levels 0 and 1.
Several studies have demonstrated that overexpression of p185 measured by immunohistochemistry is an independent prognostic factor in survival and recurrence in NSCLC. 14,15 In this study, by using the quantitative technique ELISA, which other authors had previously used in studies on recurrence, 11 we reached the same conclusions when studying survival.
Results of p185 expression measured by immunohistochemistry in relation to histologic type vary from one study to the next. 14-16
Data obtained in the different studies that measure p185 expression by immunohistochemistry are similar to those that use ELISA and demonstrate that there is no significant association between p185 expression and increasing age or more advanced stages, although higher levels of p185 are observed in well-differentiated tumors. 14-16
Marks and colleagues 17 indicated that fixation decreases the sensitivity of p185 detection by immunohistochemistry compared with the use of frozen samples. In fact, p185 expression determined by immunohistochemistry can underestimate the true incidence and level of expression in these tumors.
Díez and colleagues 11 showed that TNM staging and histologic type are the most important predictive variables in patients with p185 values below 100 U/mg, with increasing risk attributed to this marker on increasing p185 levels. Therefore in patients with p185 values above 600 U/mg, the negative influence of this protein becomes even more important than the influence of stage or histologic type.
This fact emphasizes the inadvisability of a dichotomous interpretation of the results (positive vs negative), at least when this variable is used as a prognostic marker of the disease. Other more sensitive techniques for detection and estimation of the level of p185 expression (ELISA) give additional information, especially in epidermoid tumors. 11 Therefore the relation found between levels of p185 expression and survival is different when the ELISA technique is used from when data are obtained by immunohistochemical techniques.
A recent study 15 reported overexpression of p185 in all histologic types of NSCLC and showed this to be a marker of poor prognosis in these tumors. These authors showed an additive effect of the expression of oncogenes p53 and p185, which act independently in tumorgenesis.
Expression of c-erb -B-2 has been shown to be associated with a worse outcome in cancers of the breast, 17,18 stomach, 1 ovary, 19 and lung. 15 These observations were made from p185 determinations by means of immunohistochemistry, except in the studies on breast 12 and lung cancers, 11 which used the ELISA technique and also showed an association with poor outcome.
In NSCLC some series record a lower survival in stage I patients in which p185 expression was determined by means of immunohistochemistry. 15
An association between the level of p185 expression and poor outcome was not found in any of the global series of NSCLC, which used immunohistochemistry for p185 determinations. However, an association was found in some series that focused on tumors of a specific stage 15 or histologic type. 16 This should be studied in relation to all the other factors that can be associated with or can affect survival. That is, a series of patients with lung cancer is composed of different subgroups of stages and histologic types. Possibly, as suggested previously by other authors, 20 p185 expression is not only associated with poor outcome in early stages but also with a particular histologic type. Therefore whether global survival is found to be affected by this variable would depend on the composition of the group.
Data that refer to the influence of p185 on survival suggest that this has the greatest prognostic significance in stage I tumors. In later stages a poor outcome is not dependent on p185 expression and must be related to other factors.
In our study an association was found between histologic type and p185 expression. In contrast with other authors who had used immunohistochemical techniques and reported that p185 expression was a marker of poor outcome in adenocarcinomas, our results revealed a lower survival associated with higher levels of p185 expression in patients with epidermoid tumors. These findings suggest that overexpression of p185 in epidermoid tumors is in itself of vital importance in tumoral development, whereas in the other histologic types, because there is no significant increase in p185 expression, the combination of other genetic alterations or alteration of the erb -b-2 gene possibly secondary to other genetic aberrations 21 would have to occur.
The higher values in healthy tissue from the tumoral group could occur as a result of some other process; for example, smoking induced early genetic alterations that may occur in a premalignant state. The wide range of p185 values in tumoral specimens could be explained because of the different histologic type.
One of the objectives of this study was to identify patients with a poor prognosis in the early stages to use more aggressive treatment than that currently used and therefore increase survival. Unfortunately, in vitro studies on a range of human neoplasms, including lung cancer, have shown that tumors that express p185 are more resistant to chemotherapy and radiotherapy. 22 This must be verified in patients with NSCLC. If these patients are proven to present a greater resistance to adjuvant therapies, it will be necessary to seek new alternative therapies or approaches, such as neoadjuvant chemotherapy, in stage I patients with p185 levels in the cytosol of 350 U/mg or greater. One of the lines of research that has already begun in this field deals with obtaining tumoral cell cultures. Knowledge of the in vitro sensitivity to certain drugs of cell lines derived from tumors with a probable poor evolution, determined by the presence of biologic markers of poor prognosis, could be the beginnings of a more personalized and effective therapy.
This study shows that p185 expression is an objective variable for the assessment of phenotypic aggressivity in NSCLC. In the future, a classification derived from a multiple regression study that combines the TNM classification for the anatomic description of the extent of tumoral spread and one or more variables to determine the aggressivity of the tumor could result in the generation of a specific prognostic index for each patient. Determination of p185 expression could be a good candidate for use in this multifactorial predictive model.
|
References |
|---|
|
TopAbstractIntroductionPatients and methods Results Discussion References |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
