Posts Tagged ‘Broadgreen Hospital’

Genomics of Bronchial Epithelial Dysplasia

Curator: Larry H Bernstein, MD, FCAP


C. Walker, LJ Robertson, MW Myskow, N. Pendleton & G.R. Dixon
Clatterbridge Cancer Research Trust, J K Douglas Cancer Research Laboratory, Clatterbridge Hospital, & Broadgreen Hospital, Liverpool, UK.
Br. J. Cancer (1994). 70, 297-303 expression in normal and dysplastic bronchial epithelium and in lung carcinomas 

Bronchial epithelial dysplasia is thought to be a premalignant stage in the evolution of lung cancers. Using the CM-1 polyclonal antibody, we have examined the expression of the p53 protein in a larger series of bronchial dysplasias (n = 60) than hitherto investigated. The p53 protein was detected in 14% of mild, 25% of moderate and 59% of severe dysplasias; increased p53 expression correlated with the severity of dysplasia. p53-positive dysplasias had greater PCNA indices than p53-negative dysplasias. p53 expression in dysplastic tissues was compared with that in two groups of histologically normal epithelium: 14 bronchial biopsies from non-cancer patients of which all but one were negative and 32 bronchial margins from resected carcinomas, of which 17 showed infrequent solitary cells with p53-positive nuclei in predominantly basal locations scattered throughout the epithelium. These results for resection margins were confirmed by use of a second antibody, DO-1. Sixty-nine per cent of the corresponding carcinomas were p53 positive, but in 15 cases the p53 reactivity differed from resection margins. No correlation between p53 expression and any of the clinicopathological characteristics of these tumours was found. This study supports the observation that abnormal p53 expression may be an early but not obligatory event in malignant transformation in lung.

It is now widely agreed that all lung cancers are derived from a common pluripotent stem cell capable of expressing a variety of phenotypes. Although the sequence of events in the histogenesis of lung cancer is unknown, bronchial epithelial dysplasia is thought to be a premalignant stage in the evolution of lung carcinomas. Multistep genetic changes, which include

  • activation of cellular proto-oncogenes and
  • inactivation of tumour-suppressor genes, are
    • associated with the development of human cancers and are
    • thought to accompany the morphological changes that precede malignancy.

Currently the most commonly identified genetic change in human cancers is mutation in the p53 gene, located at position 13 on the short arm of chromosome 17. This gene is

  • a tumour suppressor gene and
  • encodes a 53 kDa nuclear phosphoprotein
  • capable of binding to DNA and
  • acting as a transcriptional factor.

The wild-type p53 protein inhibits cell proliferation, and

  • loss of this activity leads to neoplastic transformation.

This protein has a

  • short cellular half-life and
  • is usually present in normal cells, under normal physiological conditions, in extremely small amounts,
  • making it undetectable by standard immunohistochemical techniques.

Many mutations of the p53 gene, principally in exons 5-8,

  • lead to a functional inactivation of the -gene and
  • a protein product unable to regulate transcription, ultimately
  • resulting in deregulation of cell growth.

Mutant p53 has

  • an extended cellular half-life
  • enabling immunohistochemical detection of
  • the accumulated mutant protein in cell nuclei.

Although not all mutations lead to protein accumulation, in many studies a correlation between the p53 protein detected immunocytochemically and p53 gene mutations has been found.
Investigation of p53 overexpression in premalignant tissues has led to the observations that

  • alterations in the p53 gene
  • arise as late events in the evolution of some cancers,
    • e.g. in gastric carcinomas, prostatic carcinomas or melanomas, whereas in others,
    • e.g. oral , gall bladder and oesophageal, malignancies, abnormal p53 expression is an early event.

In attempts to define the type and temporal sequence of somatic genetic changes that precede the onset of invasive lung cancer, recent studies have reported mutations and allelic deletions in the p53 gene in preinvasive bronchial lesions. Immunodetectable p53 has been found in a few cases of bronchial dysplasia et al., and Nuorva et al. (1993) have reported that p53 overexpression correlated with the severity of dysplasia in 17 cases of dysplastic epithelium from cancer bearing patients. Thus lesions in the p53 gene have been reported as possible early events in the development of lung cancers.

p53 score and grade of dysplasia

The system of p53 scoring used in these experiments permitted a semiquantitative comparison of the degree of p53 expression in the various tissues examined. With increasing severity of dysplasia there was not only an increase in the percentage of cases demonstrating p53 staining but also an increase in the staining intensity of positive cells and an increase in the proportions of these positive cells. Thus, higher grades of dysplasia were associated with higher p53 scores; the Spearman rank correlation coefficient for the whole table is 0.47 (P = <0.0001), and considering  just the dysplasia cases it is 0.37 (P = 0.002). Comparison of p53 expression between the various grades of dysplasia by use of p53 score  results in more significant P-values by the Mann-Whitney test than obtained with the Fisher-Irwin tests.

PCNA indices

For 39 cases of bronchial dysplasia, PCNA indices had been determined previously (Pendleton et al., 1993). p53-positive dysplasias had significantly greater PCNA indices than p53-negative dysplasias (Table IVa), indicating abnormal growth in these p53-positive biopsies

Bronchial carcinomas and resection margins

In previous studies, a series of bronchial carcinomas (Burnett et al., 1993) and their corresponding resection margins which contained histologically normal epithelium (Pendleton et al., 1993) had been collected prospectively following surgery. Using the CM-1 antibody, p53-positive nuclei were seen in 22/32 (69%) of the tumours and in histologically normal epithelium in 17/32 (53%) of the resection margins (Table V).  p53-positive cells in resection margins were predominantly basal, solitary and scattered throughout the epithelium (Figure 2a), and were less frequent than in tumour tissues (Figure 2b) or many samples of dysplastic epithelium. Many nuclei were weakly stained, but some showed a staining intensity similar to p53-positive tumour cells. Compared with dysplasias and tumour tissues the p53 scores of resection margins were low (Tables III and V), with only one case with a score of 3 and no higher scores. To confirm these results, sections of resection margins were stained with the monoclonal antibody DO-1; all of the cases positive for the CM-1 antibody were also DO-1 positive, but two cases (numbers 16 and 21) which were negative for CM-1 were clearly positive for DO-1.


During the course of the preparation of this manuscript, it has been reported that

  • the p53 protein accumulates frequently in early bronchial neoplasia.

This study differs only in that biopsies, not resected tumours, were examined and all tissues were derived from a single treatment centre.  The results of all published studies (p53 expression has so far been investigated in a combined total of 23 mild, 31 moderate and 77 severe dysplasias) yields

  • 19% of the mild,
  • 28% of the moderate and
  • 63% of the severe dysplasias
    • found to be p53 positive.

In other similar studies investigating the expression of the p53 protein in premalignant lesions of lung and other tissues,

  • results were analysed by assessment of p53 positivity.

In this study, analysis was either

    1. by comparison of p53-positive and -negative groups or
    2. by use of a p53 scoring system similar to that described by Vojtesk et al. (1993).

The advantage of this scoring system is that it allows comparison of the degree of p53 expression between tissue groups.
The p53-positive group, equivalent to the positive group in other similar studies,

  • had a p53 score of two or more.

Cells with the p53 score for intensity of 1 were clearly p53 positive and were

  • found in tumours as well as in dysplasias and normal tissues.

Unlike other studies of preinvasive lung lesions,

  • PCNA indices for many of the dysplasias in this series had been determined.

The greater PCNA indices of the p53-positive group

  • indicates that p53-positive dysplasias contain higher proportions of cells in the proliferative phase of the cell cycle;

this suggests that p53-positive dysplasias

  • may have abnormalities in their growth control mechanisms.

It is possible that alterations in the p53 gene confer a growth advantage on these cells, leading to

  • expansion of p53-positive cells as severity of dysplasia increases.

A close relation between p53 overexpression and PCNA indices has also been observed in

  • pancreatic duct cell carcinomas,
  • hepatocellular carcinomas, and
  • gastric cancers.

In this study, p53 expression in dysplastic tissues was

  • compared with two groups of histologically normal epithelium.

All but one of the first group, taken from patients who did not have cancer at the time of biopsy, were negative. Comparison of p53 expression in this group with that in

  • dysplastic bronchial biopsies
    • showed a highly significant difference between these groups.

The second group of histologically normal epithelium analysed,

  • from the resection margins of bronchial carcinomas,
    • showed p53-positive cells in a high proportion of cases,
      • indicating differences in the normal bronchial epithelium of cancer and non-cancer patients.

p53 positivity in the normal mucosa of resection margins did not result in a measurable increase in proliferation, as indicated by PCNA indices. This may suggest that the mechanism whereby the p53 protein is elevated in normal mucosa differs from that in dysplasia. Whatever the mechanism to account for these p53-positive cells in normal bronchial mucosa, it seems that their presence, even if not associated with mutation in the p53 gene, indicates abnormalities that are not reflected in the histological appearance of these cells.

The number of p53-positive tumours in the series (69% overall and 68% for non-small-cell lung cancers) agreed well

  • with the incidence of p53 positivity for lung cancers reported in some studies
  • but was higher than that found in others.

Although the number of tumours in this series was small no correlation in p53 overexpression was found with any of the clnical characteristics of these tumours. This contrasts with reports of a relationship between p53 overexpression and

  • poor prognosis and shortened survival,
  • tumour grade  or lymph node involvement and
  • a greater incidence in squamous cell carcinomas compared with other types of lung carcinoma .

This study supports the observation that abnormal p53 expression is an early but not obligatory event in the evolution of lung cancers. Immunodetection of p53 overexpression in bronchial epithelium

    • may be a useful tool in the identification of those early lesions which may progress to malignancy.
Age-standardised death rates from Trachea, bro...

Age-standardised death rates from Trachea, bronchus, lung cancers by country (per 100,000 inhabitants). (Photo credit: Wikipedia)



Read Full Post »