Dysplasia in view of the cell cycle.

Dysplasia is linked to altered tissue architecture. The lesion belongs into the diagnostic field of human pathology and is highly relevant for the clinical physician, because it breaks the criteria of hyperplasia and regeneration. Dysplasia is a precancerous disorder leading in all probability to malignant transformation if not treated. However, different descriptions do apply for dysplasia in different human tissues, and conventional pathology cannot arrive at unequivocal stringency. In contrast to the previous situation, now, dysplasia is defined by a unifying concept, which works upon cell cycle criteria. The decisive element for the proposed definition is unbalanced segregation of chromosomes and persistent genomic asymmetry through telophase, leading to aneuploid interphase nuclei. Progress of dysplasia can be estimated from the frequency of pathologic mitoses that directly measure cellular proliferation. In routine work, progress of dysplasia shall be quantified by frequency increase of aneuploidy in the increasing fraction of proliferating interphase nuclei. Thus, dysplasia is defined not only by aberrations from healthy histological architecture and normal cytological differentiation, but also by violations of the DNA standard from mitotic nuclei. The proposed classification of dysplasia measures the frequency of pathologic mitoses and the degree of genomic alterations in interphase nuclei. Both these criteria discriminate between low-grade and high-grade dysplasia and ascertain the malignant potential of a dysplastic lesion.

Genome instability in human tumorigenesis: microphotometry of interphase nuclei and pathologic mitoses reveals dysplasia.

Tumorigenesis goes with genome instability comprising point mutations and gross chromosome aberrations due to defective DNA repair mechanisms and multiple overrun of cell cycle checkpoints. Pathologic mitoses (CDFs) occur in human precancers and cancers and were detected by nuclear DNAs deviating more or less than 0.5 c from 4.0 c values. Abundant CDFs were recorded above 4.5 c threshold in lesions of the uterine cervix and the stomach. Low-grade dysplasias and well differentiated carcinomas showed 3%-10% CDFs, high-grade dysplasias 30%-44% CDFs in total divisions. Poorly differentiated cancers comprised some 50% CDFs. Most telophase CDFs showed asymmetric morphology and unbalanced DNA content in their corresponding "halves". CDFs precede DNA aneuploidy of interphase nuclei not only in precancers, but also in cancer. Chromatin bridges and lagging chromosomes suggest that unbalanced telophases are caused by somatic nondisjunction. Tumour progression from low-grade to high-grade dysplasia and cancer is characterised by recurrent shifts from 2 c to 4 c interphase nuclei and a remarkable increase in the 5 c exceeding rate. Clonal selection is the gateway in tumorigenesis for aberrant karyotypes.

Reliability and significance of DNA measurements in interphase nuclei and division figures in histological sections.

DNA contents from single cells at interphase and division were analysed in histological sections and in imprints from 73 breast cancer specimens. Fetal livers from 18 terminations of normal pregnancies provided the standard for truly mitotic prophases, metaphases and telophases. The reliability of DNA quantities from image microphotometry was improved using paraffin-embedded tissue samples from which 4, 8 and 15 microns slices were Feulgen stained. Imprinted replicas from the mirror surface of each freshly cut specimen provided matching domains and represent the crucial approach in this project. A close positive relationship was observed between interphase nuclei in 8 microns sections and their imprinted counterparts (r = 0.992; n = 73). Interphase nuclei in 4 microns sections yielded insufficient DNA contents when compared with the imprints (r = 0.815; n = 21) and with endogenous lymphocyte nuclei. This 2 cDNA standard also calibrated 232 mitotic figures to 3.91 +/- 0.01 c in 15 microns sections from fetal liver. Prophases, metaphases and telophases were slightly scattered (coefficient of variation = 0.04 each). The 0.09 c deficiency to plain 4.0 c was read as an artifact from sectioning. However, the methodical bias did not challenge the most irregular DNA distribution profiles recorded from chromosome division figures (CDFs) in 15 microns sections of breast cancers. Poorly differentiated and aggressive breast cancer (Auer type IV, Zetterberg type A) exhibited a 4.5 c exceeding rate of 82.24% from a total of 752 CDFs in 10 randomly selected cases. Well differentiated, slowly growing cancer with diploid interphase nuclei (Auer I, Zetterberg D) surprisingly showed a 4.5 c exceeding rate of 29.26% from a total of 173 mitoses and CDFs in 10 randomly selected cases. The bulk of data beyond the mitotic 4.0 c level discriminates biological bias from methodical impairment. We concluded that 8 microns sections are sufficient for human interphase nuclei, whereas a depth of 15 microns preserves intact mitoses and CDFs.

The DNA content of chromosome division figures and interphase nuclei classifies ulcerative colitis.

Long-standing ulcerative colitis is considered to be a precancerous condition. Therefore, a practical and reliable method is required for monitoring the progress of the disease. Liberation of the S-phase from karyokinesis occurs in DNA amplification and endoreplication, producing nuclei with more than 4 c DNA. The amount of Feulgen DNA was quantified with an image microphotometer in 8 microns sections for interphase nuclei and in 15 microns sections for chromosome division figures (CDFs). Development of ulcerative colitis was investigated in low grade dysplasia (n = 93 cases; score 3-7) and high grade dysplasia (n = 22; score 8-10). Bacterial colitis (n = 34) and invasive adenocarcinoma (n = 26) provided a basis for data interpretation in dysplasia. Lymphocyte nuclei served as an internal DNA standard. CDFs represent a novel type of aberrant 'mitoses'; they are different from and much more frequent than figures with multipolar spindles. Endoreplication began with low grade dysplasia in interphase nuclei as well as with CDFs; it was fully established in high grade dysplasia and carcinoma. Endoreplicated interphase nuclei and CDFs represent an early morphological mosaic of genomic instability. Both characteristics support a reproducible two-level classification of low and high grade dysplasia in ulcerative colitis.

Chromosome division figures reveal genomic instability in tumorigenesis of human colon mucosa.

A variety of chromosomal gains and losses has been detected with comparative genomic hybridization during tumorigenesis in the colon mucosa. The aim of this investigation was to corroborate increasing genomic instability and to elucidate those lesions in which the record from comparative genomic hybridization has remained unexpectedly negative. Replicate paraffin-embedded samples were investigated in detail using image microphotometry. Crucial to the recent approach was the fact that the histological compartments were exactly matched and that the single-cell measurements were highly accurate (CV at 0.05). Feulgen DNA was quantified in interphase nuclei and chromosome division figures, which were found in all cases of high-grade dysplasia and, with increased frequency, of colon carcinoma. The genomic imbalance in chromosome division figures was quantified by the sensitive 4.5 c exceeding rate (where c is the haploid genome equivalent), which was also positive in cases with a negative record from comparative genomic hybridization. The DNA content of chromosome division figures was measured with a mean 4.5 c exceeding rate of 25.8 +/- 4.4% standard error in 12 cases of high-grade dysplasia and of 62.1 +/- 7.1% in colon carcinoma (16 cases). The chromosome division figures were considered to be the first morphological manifestation of genomic instability attending precancerous conditions in the colon. Telophase-like chromosome division figures with unequal amounts of DNA in their hemispheres revealed gross somatic mutations before clonal selection.

Atypical mitoses in lesions of the oral mucosa: a new interpretation of their impact upon tumorigenesis.

With regard to the classification of precancerous lesions, this is the first approach to compare modes of the cell cycle in benign lesions and neoplasia in oral mucosa. The main objective of this study was to find parameters that indicate possible progress in tumorigenesis. In this respect, histomorphological changes were compared with the DNA content of individually identified nuclei in interphase and division. Thus, hyperplastic tissue was discriminated from neoplastic tissue by investigating pyogenic granulomas (27 cases), epulides (17), simple forms of ulcers (12), low-grade dysplasias (14), high-grade dysplasias (15) and carcinomas (41). After diagnosis, Feulgen DNA was quantified from interphase nuclei and chromosome division figures (CDFs) in 8- and 15-micron sections. Interphase nuclei from cases of hyperplasia were constantly found in the range of mitotic amounts, i.e. 2 c-4 c DNA. A 4 c value was therefore recorded for mitotic figures. Measurements in this type of lesion were carried out with a high degree of accuracy (CV at 0.05). However, cases of neoplasia showed CDFs that were frequently beyond the mitotic range (> 4.8 c) in morphologically identified prophases, metaphases and both hemispheres of telophases. The distribution profile of interphase nuclei was characterised by DNA aneuploidy, with a 5 c exceeding rate (ER) > 5%. This biological variability in the amount of nuclear DNA was reflected by CV > 0.20. The quantitative results corroborated the morphological two-level classification of premalignant lesions caused by endoreplication associated with neoplasia. Thus, regular mitotic replication appeared to be progressively substituted by genome multiplication.

Proliferation and DNA aneuploidy in mild dysplasia imply early steps of cervical carcinogenesis.

This project was focused on cellular proliferation relative to the onset of endoreplication and effects of DNA aneuploidy during carcinogenesis in cervical mucosa. Proliferation was monitored with MIB1 antibody, whereas nuclear DNA content was quantified using an image processing microphotometer. For the later procedure, 8 microm sections were of adequate depth with interphases, and lymphocyte nuclei provided an internal standard of the diploid (2c) DNA content. Results from 95 cervical biopsies displaying different types of dysplasia and carcinoma were supplemented with those of cervical smears from 319 cases. The later specimens had been selected from about 30000 consecutive cases in 1993/94. MIB1-traced proliferation was found in the second cell layer, whereas the bulk of basal cells remained quiescent in normal mucosa. However, predominant MIB1 immunoreactivity was observed with endoreplicated nuclei. A critical 30% of the cases exhibited DNA aneuploidy already in mild dysplasia, which was found in cytological smears and histological sections. The nuclear DNA content of basal cells increased progressively by endoreplication corresponding to the degree of dysplasia. Cases of carcinoma in situ displayed some 18% of non-proliferating diploid cells despite overwhelming endoreplication and DNA aneuploidy. High MIB1 levels combined with DNA aneuploidy unambiguously indicate the beginning of cervical carcinogenesis. The limits of the Bethesda system were discussed.

Imbalance of morphologically addressed telophases reflects interphase DNA aneuploidy in tumorigenesis.

Chromosome division figures (CDFs) are quantitatively different from normal mitoses and represent a novel cytogenetic phenomenon. This investigation was focused on morphologically addressed bipolar telophases in histologically defined human biopsies and in the tumour breast cell-line MDA231. Single cell nuclei were recorded by image microphotometry on inflamed and premalignant lesions of skin (49 cases), oral mucosa (43) and colon mucosa (46). DNA content and replication status were analysed in interphase nuclei as well as in mitoses and in CDFs. In contrast to inflamed lesions, premalignancies were characterised by pronounced endoreplication, when the rate exceeding 5 c was > or = 10% in interphase nuclei. CDFs from the corresponding lesions showed an aberrant DNA content beyond 5 c even more frequently. DNA profiles of metaphases and telophases resembled those of prophases. Therefore, the DNA content of corresponding telophase hemispheres was measured. Severe differences averaged 0.3 c in MDA231 and up to 0.5 c in premalignant lesions. The mean difference between two corresponding hemispheres was 0.39 +/- 0.09 c in Bowenoid keratosis (n = 31), 0.40 +/- 0.08 c in high-grade dysplasia of oral mucosa (n = 16) and 0.21 +/- 0.03 c in high-grade dysplasia of colon adenoma (n = 65 telophases). As a control, the telophase difference was only 0.07 +/- 0.02 c (n = 23) in foetal liver and 0.06 +/- 0.01 c in 24 amnion cells. Thus, genomic instability and, in consequence, genomic imbalance can best be quantified from the DNA profiles of telophase CDFs and from the various DNA amounts in their hemispheres. A strong selection against telophases was observed in neoplasias developing DNA aneuploidy. Those aberrant telophases which escape selection are thought to enhance tumour progression.

The relationship between proliferating cell nuclear antigen (PCNA), nuclear DNA content and mutant p53 during genesis of cervical carcinoma.

Proliferating cell nuclear antigen (PCNA), nuclear DNA content and mutant p53 overexpression were studied by means of image cytometry and immunohistochemistry respectively in normal mucosa (n = 10), in mild (n = 16), moderate (n = 9) and severe (n = 17) atypical lesions, as well as in squamous cell carcinomas (n = 36) of the cervix uteri. The results show that increasing histopathological atypia in the cervical mucosa was correlated to an initial increase of PCNA followed by distinct aneuploidy and p53 overexpression. The data are suggested to contribute to a better understanding of the genesis of cervical carcinoma, and to indicate that the coexistence of both distinct aneuploidy and p53 immunoreactivity can be used to decide if a cell population is neoplastic, whereas the absence of p53 overexpression does not necessarily exclude neoplasia. This diagnostic procedure can be suggested to improve early detection of intraepithelial squamous neoplasia.

DNA ploidy in human colorectal adenomas.

The DNA content of 101 colorectal adenomas of various histologic types resected from 83 patients was determined by image cytometric measurements in order to investigate if a correlation between DNA ploidy and particular histomorphologic features exists and if DNA measurements can be of additional diagnostic value. Overall, 67 of 101 (66%) adenomas showed an aneuploid DNA distribution pattern, including 8 of 19 (42.1%) mildly atypical, 32 of 48 (66.7%) moderately atypical and 27 of 34 (79.4%) severely atypical adenomas. Correlating DNA content with the histologic type, 17 of 42 (40.5%) tubular, 28 of 37 (75.7%) tubulovillous and all 22 villous adenomas exhibited aneuploid DNA histograms. Aneuploidy was also observed more frequently in larger adenomas. The results show a good correlation between tumor size, histomorphologic features and DNA content. The most remarkable observation is that as many as 42% of the adenomas histomorphologically considered mildly atypical exhibited aneuploidy. Since aneuploidy has been demonstrated to indicate premalignant or malignant cellular alterations, DNA image cytometry is suggested for providing valuable additional information on the diagnosis of colorectal adenomas.

The relationship between aneuploidy and p53 overexpression during genesis of colorectal adenocarcinoma.

This paper describes the investigation of nuclear DNA content and p53 immunoreactivity in normal mucosa (n = 25), mildly (n = 15), moderately (n = 28) and severely atypical (n = 22) colorectal adenomas and in colorectal adenocarcinomas (n = 116). Twenty-seven per cent of the mildly atypical, 43% of the moderately, 77% of the severely atypical adenomas and 91% of the colorectal carcinomas were distinctly aneuploid. In the aneuploid lesions p53 immunoreactivity was not observed in mildly atypical adenomas, whereas 17% of the moderately atypical, 24% of the severely atypical adenomas and 66% of the adenocarcinomas were p53 positive. None of the diploid lesions were p53 immunoreactive. These data are interpreted to indicate that genomic instability as reflected by crude aneuploidy occurs early during genesis of colorectal carcinoma and represents a high risk factor for p53-gene mutation.

DNA content and PCNA immunoreactivity in oral precancerous and cancerous lesions.

Thirty-three dysplastic lesions showing varying degrees of atypia located in the oral cavity and 83 squamous cell carcinomas located in the oral cavity and tongue (n = 56) or in the lips (n = 27) were analysed by means of proliferating cell nuclear antigen (PCNA) immunoreactivity and image cytometric DNA measurements. The results show that in dysplastic lesions increasing cellular atypia correlated to elevated proliferative activity and aneuploidy occurring in the basal cell layers. In highly differentiated squamous carcinomas increased PCNA immunoreactivity and aneuploidy was preferentially observed focally (grade 1 tumours) or in the invasive zones (grade 2 tumours). In contrast, more poorly differentiated carcinomas (grade 3 and 4 tumours) showed strongly elevated proliferative activity and aneuploidy throughout the entire tumour mass.

DNA ploidy in human colorectal adenocarcinomas.

Image cytometric determination of DNA content was performed on 207 colorectal adenocarcinomas surgically resected from 205 patients. Of these 207 tumors, 193 (93.2%) showed an aneuploid DNA distribution pattern, whereas the remaining 14 (6.8%) exhibited a DNA pattern corresponding to that found in proliferating diploid populations. The DNA content of 20 of these colorectal carcinomas was also studied by flow cytometric measurements on deparaffinized, disintegrated material. Flow cytometric measurement detected aneuploidy in only 9 of these 20 (45%) tumors, whereas 17 of the same 20 (85%) carcinomas showed aneuploidy when using image cytometric measurement on imprints. These results suggest that colorectal carcinomas are either aneuploid (vast majority) or proliferating diploid and that image cytometric DNA measurement on histologically controlled material is superior to flow cytometric measurement.