Cytomorphologic and DNA cytometric features of hepatocellular carcinoma in fine needle aspirates

OBJECTIVE

To classify hepatocellular carcinoma according to DNA ploidy patterns and to evaluate distinct cytomorphologic features of hepatocellular carcinoma that correlate with DNA ploidy patterns.

STUDY DESIGN

Fine needle aspiration smears of 36 histologically proven hepatocellular carcinomas were performed for DNA measurement by image analysis after Feulgen restaining of the specimens. Nuclear features-prominent nucleoli, nuclear cleavage, nuclear area and nuclear/cytoplasmic ratio-were correlated with the DNA ploidy patterns.

RESULTS

Of the 36 cases, 14 were either diploid (n = 7) or polyploid (n = 7), 19 tumors had a single aneuploid stemline, 2 cases had multiple stemlines, and 1 case had no discernible stemline. A preponderance of prominent nucleoli was seen in 7/7 diploid tumors (2c), 6/7 polyploid tumors (4c, 8c) and 6/8 aneuploid tumors (> 4c). Conspicuous nuclear cleavage in a high number of tumor cells was present substantially in tumors with large nuclear areas (4c, > 4c).

CONCLUSION

Most hepatocellular carcinoma studied had a distinct stemline so that the tumors could be designated DNA diploid, polyploid or aneuploid. The prevalence of prominent nucleoli and nuclear cleavage was a distinguishing cytologic feature that could predict DNA ploidy patterns. No special association of the nuclear/cytoplasmic ratio with any of the ploidy groups was noted.

Canine neuroendocrine tumors of the pancreas: a study using image analysis techniques for the discrimination of metastatic versus nonmetastatic tumors

Canine pancreatic neuroendocrine tumors were studied using different image analysis techniques (nuclear image histometry, analysis of argyrophilic proteins of nucleolar organizer regions, determination of the mouse anti-Ki 67 antigen proliferation index, and DNA densitometry) to correlate their biological behavior with objective phenotypic markers. The methods were compared to determine the best method for distinguishing between metastatic and nonmetastatic tumors. Discrimination between the two types of tumor was possible using nuclear image histometry in combination with morphometric analysis of argyrophilic proteins of nucleolar organizer regions. In contrast, the mouse anti-Ki 67 antigen proliferation index, DNA measurement, and immunohistochemical parameters revealed no significant difference between the two types of tumors.

Segmentation of sperms using the strategic Hough transform

A new method based on the Hough transform has been used for the quantitative estimation of the morphological characteristics of the sperm. Images of the sperms were acquired into the digital format using the optical microscope, CCD camera, and frame grabber. For each sperm in the image, the region of interest for the segmentation of the sperm head was selected using the density difference between the sperm head and background. The boundary of the sperm head was approximated with an ellipse and represented by five parameters. These five parameters representing an ellipse have been investigated by applying the Hough transform strategically. During the searching procedure using the Hough transform, we restricted the searching volume minimally and moved the next searching volume to the direction showing the steepest gradient of the estimation error. Also, we increased the parameter resolution from the three pixel distance to the one pixel distance. For the tested images, head boundaries of the sperms were estimated with an error of 1/2 pixel distance, and morphological parameters were calculated by the equations without further image processing. This new method separates sperm head based on the geometrical characteristic of the sperm and calculates morphological parameters simply from the derived boundary.

Sample preparation from paraffin-embedded tissue specimens for laser scanning cytometric DNA analysis

We have developed a simple, rapid method for isolating cells from a block of formalin-fixed, paraffin-embedded tissue specimen for laser scanning cytometric (LSC) DNA analysis by using a grater. The scraping-like tissue samples were obtained by grating a paraffin-embedded tissue block. The grated samples were collected, put into a small plastic tube, and deparaffinized with xylene. Subsequently, the samples were immersed in 100% ethanol to remove the xylene. After using a syringe with a 26-gauge needle and filtering through 40-microm nylon mesh, the cells suspended in ethanol were dropped directly onto a glass slide. As a result, isolated cells adhered tightly to the glass slide. The slides mounted with isolated cells were treated with 0.1% pepsin in 0.1 N HCl for 1 h at 37 degrees C and then 0.1% RNase for 10 min at room temperature. The slides were dipped in propidium iodide (25 microg/ml) to stain DNA and sealed with nail varnish. The coefficients of variation for histograms were small enough to detect an aneuploid peak close to the diploid peak.

Symmetric chromosomal order in leukocytes indicated by DNA image cytometry and fluorescence in situ hybridization

OBJECTIVE

To study if the symmetry observed in the nuclear structure of polymorphonuclear cells (PNC) could be verified functionally by determining DNA and chromosome distributions using image cytometry (ICM) and fluorescence in situ hybridization (FISH) STUDY DESIGN: In murine and human leukocytes the proportion of PNCs with apparently bilaterally symmetric nuclei was determined. By ICM, total nuclear DNA and DNA content in each half of these nuclei were measured. For chromosomal localization the FISH technique was applied.

RESULTS

About 13% of PNC nuclei appeared to be bilaterally symmetric. ICM of Feulgen DNA revealed that the PNC had a 2N value, similar to mononuclear cells. The DNA content in each of the two symmetric parts of the nucleus was exactly the same. Similarly, FISH signals specific to chromosomes 1 and 2 were distributed bilaterally on the two symmetric parts.

CONCLUSION

The distribution of DNA and chromosomes studied support the hypothesis of bilateral symmetry of PNC nuclei, suggesting haploid separation. This specific distribution may be valuable for the study of mechanisms leading to DNA and chromosome damage by exposures to chemicals or radiation and may have far-reaching significance in genetics regarding imprinting, meiotic pairing and other subjects.

Quantitative measurement of estrogen and progesterone receptors in breast lumps using image cytometry

OBJECTIVE

To evaluate image cytometry for detecting estrogen (ER) and progesterone receptors (PR) in breast lumps quantitatively.

STUDY DESIGN

Fifty consecutive breast lumps were analyzed for ER and PR using the avidin-biotin complex on tissue sections.

RESULTS

Mean ER-positive nuclear area percentage (PA%) for 25 invasive ductal carcinomas was 47.81% and for PR was 57.83%. Mean ER PA% was 40.78% and PR 58.91% for seven invasive lobular carcinomas. There were two papillary carcinomas and one in situ ductal carcinoma. Fourteen of 15 benign breast lumps were benign breast lesions. Mean ER PA% was 65.95% and PR 73.43% for 10 fibroadenomas and for 4 fibrocystic lesions was 51.57% and 44.13%, respectively.

CONCLUSION

Quantitative detection of ER and PR can be achieved by image cytometry. We aim in the future to assess this method in prognostic studies and the preoperative cytology of breast cancers.

Three-dimensional DNA image cytometry by confocal scanning laser microscopy in thick tissue blocks of prostatic lesions

DNA ploidy provides important information for the evaluation of the prognosis of prostate cancer. For the purpose of DNA cytometry and nuclear measurements, we developed an image processing system for the acquisition and processing of three-dimensional (3D) images based on confocal scanning laser microscopy (CSLM). The advantage of the CSLM is the preservation of the tissue architecture and the possibility of multilabeling. It is possible to determine both individual nuclear features and cellular features and the degree of the spatial heterogeneity of several markers. Special attention was paid to the development of the automatic method for the 3D segmentation of cell nuclei. Thick tissue slides (100 microm), stained for DNA with chromomycin A3, from 4 patients (with benign hyperplasia, prostatic intraepithelial neoplasia (PIN), and well-and poorly-differentiated adenocarcinoma of the prostate), were studied in order to test the practicability of the developed methodology. DNA histograms showed a single peak in the diploid range for the hyperplasia and PIN cases. For the case of well-differentiated carcinoma, 2 peaks were observed, 1 in the diploid range and I in the tetraploid range. The case of poorly-differentiated carcinoma was characterized by an aneuploid distribution. For the cases of PIN and carcinomas, we observed a considerable variation of the volume of nuclei.

Exploring the possibility of DNA ploidy measurements in tissue sections using liver as a model

OBJECTIVE

To evaluate DNA image cytometry (ICM) in tissue sections, human and rat liver was used as a model. These tissues have distinct DNA ploidy histograms of mainly diploid (human) and predominantly tetraploid (rat) plus octaploid nuclei.

STUDY DESIGN

Section thickness varied from 4 to 12 microns in rat liver and 4 to 10 microns in human liver, in steps of 2 microns. The measurements were made using a commercially available ICM system, strictly following a measurement protocol.

RESULTS

Four-micrometer-thick sections gave very broad histograms. Six-micrometer-thick sections gave better results, but with increasing DNA PLOIDY (> 4c), modal peaks were broad and increasingly shifted to the left. Sections of 8-10 microns gave linear results of the 2c, 4c and 8c peaks and also the narrowest peaks. The 12-micron sections of rat liver were more difficult to measure, and the results were not improved.

CONCLUSION

The 8-micron sections gave the best results and were easiest to measure. These results are promising, but since liver is a fairly simple model, further studies are required to elucidate whether DNA ploidy measurements can be performed reliably on tissue sections of human tumors.

Primary malignant melanoma of the skin. Relationships of nuclear DNA content, nuclear morphometric variables, Clark level and tumor thickness

OBJECTIVE

To investigate how nuclear morphometric variables, tumor thickness (measured according to Breslow), invasion depth (classified according to Clark), nuclear DNA content and type of DNA histogram are associated with each other in primary malignant melanomas of the skin.

STUDY DESIGN

Image analysis DNA cytometry and nuclear morphometry were performed on 85 primary skin melanomas. The relationships of size, sphericity and DNA content of melanoma cell nuclei; melanoma thickness; and Clark level were analyzed in detail. The effect of melanin bleaching on DNA cytometry results was studied.

RESULTS

Melanoma thickness correlated with nuclear size in aneuploid, but not diploid, melanomas. The prevalence of aneuploidy did not increase with tumor thickness. In aneuploid melanomas the proportion of cells with higher-than-diploid and higher-than-tetraploid DNA content increased with tumor size.

CONCLUSION

Aneuploidy is as common in thin as in thick melanomas. Genetic instability in aneuploid melanomas correlates with melanoma thickness. This correlation in aneuploid melanomas partially explains the correlation between nuclear size and melanoma thickness. In diploid melanomas no correlation was observed between nuclear size and melanoma thickness. DNA cytometry is a valuable tool for studies on the background of phenotypic changes in skin melanomas.

Prognostic value from DNA quantification by static cytometry in breast cancer

OBJECTIVE

To determine if DNA quantification on fine needle aspiration (FNA) has predictive value in breast cancer.

STUDY DESIGN

Forty-nine patients with breast cancer were selected for this study because of their common characteristics. The smears were studied using image cytometry.

RESULTS

With the type of histogram and value of entropy, two large groups, with high and low degrees of malignancy, were obtained. Survival was utilized as a variable of interest. Kaplan-Meier survival curves for both groups were formulated, and the results were supported with statistical data.

CONCLUSION

The survival differences between both groups were statistically significant (P < .001), thus demonstrating the predictive value of DNA quantification.

Methods for automatic multiparameter analysis of fluorescence in situ hybridized specimens with a laser scanning cytometer

Multiparameter laser scanning cytometry has been applied to the automatic counting of probe spots and the simultaneous measurement of cellular DNA for fluorescence in situ hybridization (FISH) prepared specimens counterstained with propidium iodide. Relatively low resolution imaging, highly variable probe fluorescence, spectral overlap of probe with counterstain fluorescence, and autofluorescence required the development of an image processing method to detect and isolate FISH probe spots. Inability to properly apportion detected probe spots because of overlapping probe spot images in the same cell required development of a method to eliminate cell data whenever spots in that cell could not be reliably isolated. Laser scanning cytometry incorporating these methods to determine per cell probe spot count and DNA is demonstrated on tissue cultures and peripheral blood cells using different centromeric FISH probes with either FITC or Spectrum Green labeling.

Topographical analysis of spatial patterns generated by a cellular automaton model of the proliferation of a cancer cell line in vitro

A well-suited model to simulate cellular population dynamics is the two-dimensional cellular automaton model, which consists of a lattice of sites, the value ai,j of each site being updated in discrete time steps according to an identical deterministic rule depending on a neighbourhood of sites around it. A cellular automaton is described which mimics cell population proliferation by replacing the site values by the age and the cycle phase of cells. The model takes into account the size of the cells. It is used to simulate the proliferation of the human breast cancer cell line MCF-7 and the results of the simulation are compared with experimental data obtained from a light microscopic image analysis of the proliferation process. The initial configuration of the cellular automaton is obtained from the discretization of the results of the initial stage of the image processing. After each day of proliferation the pattern obtained from the simulation is compared to the experimental result of the corresponding image analysis. The comparison is made from a topographical point of view through the concept of the minimal spanning tree graph. The agreement between experiment and model is a good starting point to complex models such as cell proliferation under growth effectors or drugs.

Examination of DNA-ploidy in melanocytic nevi cells using video-imaging cytometry

DNA content in 61 different types of melanocytic nevi was determined by video-imaging cytometry. The nevi were selected for this study because of melanoma risk associated with each nevus, which is 20-50% according to different authors. DNA ploidy was detected in paraffin embedded and fresh tissue material of each patient and the results were comparable. The sample preparation process and video-imaging method were presented in this study, 47 (77.0%) lesions exhibited diploid cell populations, 13 (21.3%) aneuploid and 1 (1.7%) tetraploid cell population. A significant correlation was observed between DNA ploidy measured in fresh tissue and retrospective material. The results indicate the presence of abnormal DNA content in some of the lesions.

Detection of malignancy associated changes in cervical cell nuclei using feed-forward neural networks

Normal cells in the presence of a precancerous lesion undergo subtle changes of their DNA distribution when observed by visible microscopy. These changes have been termed Malignancy Associated Changes (MACs). Using statistical models such as neural networks and discriminant functions it is possible to design classifiers that can separate these objects from truly normal cells. The correct classification rate using feed-forward neural networks is compared to linear discriminant analysis when applied to detecting MACs. Classifiers were designed using 53 nuclear features calculated from images for each of 25,360 normal appearing cells taken from 344 slides diagnosed as normal or containing severe dysplasia. A linear discriminant function achieved a correct classification rate of 61.6% on the test data while neural networks scored as high as 72.5% on a cell-by-cell basis. The cell classifiers were applied to a library of 93,494 cells from 395 slides, and the results were jackknifed using a single slide feature. The discriminant function achieved a correct classification rate of 67.6% while the neural networks managed as high as 76.2%.

A novel image cytometric method for quantitation of immunohistochemical staining of cytoplasmic antigens

Evaluation of molecular markers by immunohistochemical labelling of tissue sections has traditionally been performed by qualitative assessment by trained pathologists. For those markers with a staining component present outside of the nucleus, there has been no image histometric method available to reliably and consistently define cell interfaces within the tissue. We present a new method of approximating cellular boundaries to define cellular regions within which quantitative measurements of staining intensity may be made. The method is based upon Voronoi tessellation of a defined region of interest (ROI), and requires only the position of the nuclear centroids within the ROI. Here we describe the VORSTAIN software which has been developed based on the Oncometrics CytoSavant Automated Image Cytometry System. To demonstrate this technique, human breast cancer sections immunohistochemically stained for bcl-2 protein and counter-stained with nuclear methyl green stain were evaluated. Intra-observer variation in the measured values was between 1.5-2.6% and inter-observer variation was between 1.8-4.4%. The primary source of variability was due to difficulties in interpreting the exact position of the nuclear centroids. Analysis of mean staining densities for each slide correlated well with subjective scoring performed by two independent pathologists. Using VORSTAIN, significant variation of staining intensities between regions within the same slide was measured for some sections, indicating a large degree of heterogeneity within the tumours. The ability to accurately quantitate the degree of heterogeneity of molecular marker expression within tumours may be a valuable tool in prognostication.

Quality assurance in DNA image analysis on diploid cells

DNA image analysis is presently performed in many laboratories. Before general extrapolation of results between different laboratories, validation has to be performed for interlaboratory studies on DNA image analysis. The aim of this study was to establish the performance of different DNA image analysis instruments when measuring different diploid cells. On three separate parts of the same object slide, human liver cells, white blood cells, and imprints of a breast fibroadenoma were sampled. In this quality assurance interlaboratory study, 13 laboratories participated voluntarily. Two slides were sent to each participating laboratory: one Feulgen and one unstained to be stained according the participating laboratory in-house procedure. The features integrated optical density (IOD) and object AREA were recorded for each nucleus. For calculation of the results, the average IOD value of liver diploid cells was set at 2c. A striking difference was observed between the different presumed diploid cell types, from almost 1c to 3c. This variation was not dependent on central or in-house staining. Although in-house calibration was performed for each image analysis system, a surprisingly large variation existed in the reported object AREA, irrespective of the cell type. These results clearly demonstrate that measurements performed in one laboratory cannot be extrapolated to the outcome of others and support the need for standardization. The use of external control cells works well for comparison of instruments. In conclusion, in DNA image analysis quality control is necessary, standardization is obligatory, and the use of an internal control for determination of the diploid peak in a histogram of patient samples is recommended.

Diagnosis and treatment of superficial transitional cell carcinoma of the bladder: facts and perspectives

The detection of potentially highly curable low-grade bladder cancers by noninvasive techniques remains an unsolved problem. Conventional cytology detects such tumors with 50% sensitivity, the addition of DNA image analysis cytometry in patients with previous transitional cell superficial bladder carcinoma increases the chance to detect low-grade tumors. The specificity and sensitivity of detection of low-grade bladder tumors can be further improved by quantitative fluorescence image analysis using antibodies directed against tumor-associated antigens. The chance of subsequent tumor following initial complete resection is significantly lowered compared to resection alone by either intravesical adjuvant chemotherapy or immunotherapy. BCG is superior to chemotherapy in patients with high risk for recurrence, adjuvant intravesical chemotherapy does not prevent progression to > T1 when compared with resection alone, BCG may delay progression and the need for cystectomy in high-risk patients (T1, Cis or Cis associated with papillary Ta/T1, > 3 recurrences/year, grade 3 tumors). Oral medication with bropirimine (an interferon inducer) induced complete response in patients with carcinoma in situ, oral Lactobacillus casei prolonged the recurrence-free interval to 1.8 times that in the control group. New compounds for adjuvant treatment of superficial bladder tumors are best first tested on their ablative effects (marker lesion, Cis) prior to be used for prophylaxis.

Signal amplification of FISH for automated detection using image cytometry

The purpose of this study was to improve the detection of FISH signals, in order that spot counting by a fully automated image cytometer be comparable to that obtained visually under the microscope. Two systems of spot scoring, visual and automated counting, were investigated in parallel on stimulated human lymphocytes with FISH using a biotinylated centromeric probe for chromosome 3. Signal characteristics were first analyzed on images recorded with a coupled charge device (CCD) camera. Number of spots per nucleus were scored visually on these recorded images versus automatically with a DISCOVERY image analyzer. Several fluochromes, amplification and pretreatments were tested. Our results for both visual and automated scoring show that the tyramide amplification system (TSA) gives the best amplification of signal if pepsin treatment is applied prior to FISH. Accuracy of the automated scoring, however, remained low (58% of nuclei containing two spots) compared to the visual scoring because of the high intranuclear variation between FISH spots.

Automatic volumetric capillary cytometry for counting white cells in white cell-reduced plateletpheresis components

BACKGROUND

As the benefits of white cell (WBC)-reduced blood components become increasingly apparent, the need has arisen for a simple, automated WBC-counting technique that is sensitive to low WBC concentrations. Automated volumetric capillary cytometry was evaluated for its ability to quantify residual WBCs in WBC-reduced plateletpheresis components.

STUDY DESIGN AND METHODS

The volumetric capillary cytometry system evaluated uses a laser to excite fluorescent dye-labeled nucleated cells. The number of nucleated cells per microliter is reported. Four studies were performed: linearity, precision of results near the value of 5 x 10(6) WBCs per unit, the limit of detection, and correlation to the Nageotte manual counting method.

RESULTS

Assay values correlated to expected values (range, 0-125 WBC/microliter) with an r2 > 0.99. In the range of 5 x 10(5) WBCs per unit the CV was 8.5 percent, and concentration differences of 0.15 log10 were detectable. The limit of detection was 1.0 WBCs per microliter (95% upper confidence limit). The assay correlated to the Nageotte method with an r2 of 0.98, slope of 1.0, and y-intercept of 2.0 WBCs per microliter. Assay results were 10 to 15 percent higher than Nageotte results, in samples with values near 5 x 10(6) WBCs per unit. Technician time per sample was 2 to 3 minutes.

CONCLUSION

Volumetric capillary cytometry is precise and sensitive to small differences in WBC concentration in the range of clinical interest. The device provides an efficient new method for quality assurance and control of WBC-reduced plateletpheresis products.

Multiparameter absorption measurements in automated microscopy. Simultaneous quantitative determination of DNA and nuclear antigen

OBJECTIVE

To test a dual DNA-nuclear antigen staining method for multiparameter absorption image analysis.

STUDY DESIGN

MCF 7 cells, grown on glass slides, served as a model to test the staining technique. For DNA, Feulgen-based CAS quantitative DNA staining, and for nuclear antigen, alkaline phosphatase-based immunocytochemical staining with CAS Red as the chromogen, were used. MIB-1, estrogen and progesterone receptors were used as examples of nuclear antigen staining. Measurements were performed with the DISCOVERY image analyzer.

RESULTS

Scatterplots, in which the nuclear antigen content was plotted against the DNA content, were obtained. Immunostain-positive and -negative populations could be discriminated. These cells were visualized in image galleries. The DNA histograms of the positive and negative cells showed no change in coefficient of variation or integrated optical density ratio of the G0, G1 and G2 + M peaks as compared to single DNA staining. The intensity of the immunostain increased as compared to the single immunostaining result.

CONCLUSION

This staining technique allows the simultaneous accurate measurement of costained DNA and antigen within the same nucleus. This opens the possibility for studies in which nuclear antigen expression is monitored during the cell cycle or in cells of different ploidy classes. Identified cells can also be visualized by presentation in an image gallery or by relocation on the slide. This can support the analysis of clinical samples, where cytometric data can be correlated with and confirmed by visual diagnosis.

DNA measurement of overlapping cell nuclei in thick tissue sections

The paper describes an improved image analysis procedure for measuring the DNA content of cell nuclei in thick sections of liver tissue by absorption densitometry. Whereas previous methods only permitted the analysis of isolated nuclei, the new technique enables both isolated and overlapping nuclei to be measured. A 3D segmentation procedure determines whether each object is an isolated nucleus or a pair of overlapping nuclei; in the latter case the combined optical density is redistributed to the individual nuclei. A selection procedure ensures that only complete nuclei are measured. The method has been tested on specially-prepared Feulgen-stained 20micro sections of normal liver tissue. The overall distribution of the nuclear DNA measurements shows well-defined diploid and tetraploid peaks, with coefficient of variations of less than 10%. Similar distributions were obtained from both the isolated nuclei and overlapped nuclei sub-populations.

DNA-ploidy studies on cytological preparations from breast cancers by image analysis: comparison with Feulgen staining performed in destained Papanicolaou slides

Image cytometric analysis of DNA ploidy status of 40 fresh primary breast carcinomas was performed in a comparative study on cytologic preparations stained directly by the Feulgen method and smears from the same tumors Feulgen-stained after Papanicolaou staining. There was a 0.82 overall concordance in diploid and aneuploid tumor classification and a good correlation (r = 0.736, P < 0.001) between the DNA indices determined by the two methods. Discordances are probably due to tumor heterogeneity. This study demonstrates that image cytometric DNA analysis on previously routine-stained cytologic preparations is feasible and reliable for pretreatment DNA analysis and suggests potential applications in retrospective studies with cytologic material from archives.

Image cytometry of early nuclear events during apoptosis induced by camptothecin in HL-60 leukemic cells

The nuclear morphological alterations occurring during apoptosis induced in human HL-60 cells by camptothecin were analyzed quantitatively by image cytometry. Two separate phases were identified during the apoptotic process. The first phase, observed between 0-2 h of incubation, consisted in the appearance of cells with an apparently decondensed chromatin. This phenomenon was blocked by the inhibitors of DNA fragmentation, TLCK and H7. In contrast, staurosporine and cytochalasin B, which inhibit apoptosis without any effect on DNA fragmentation in this system, did not prevent this morphological change. The second phase, observed after 3 h of culture, corresponded to the appearance of cells with very homogeneous and condensed chromatin. This phenomenon correlated with the detection of typical apoptotic cells with fragmented nuclei and was inhibited by all drugs (TLCK, H7, staurosporine, and cytochalasin B). These observations suggest that image cytometry allows the detection of subvisual microscopic events within the first hour after the induction of an apoptotic process and that the dissection of this process into several different phases might be associated with DNA fragmentation.

Testing the value of a measurement protocol for DNA image cytometry and comparing different cytologic preparations

OBJECTIVE

To analyze the influence of measuring parameters on the quality of DNA histograms, and to compare the results of DNA image cytometry (ICM) and flow cytometry (FCM).

STUDY DESIGN

Material from 20 breast cancer cases was analyzed. The protocol for ICM analyzed (1) measuring system and image analysis parameters, (2) the quality of fixation, and (3) the presence of cytoplasm around nuclei. In Feulgen-stained slides (both fresh imprints and cytocentrifuge preparations) at least 200 nuclei were examined using a commercially available ICM system. DNA FCM on disaggregates was performed, also.

RESULTS

Strict adherence to the protocol gave reliable DNA ploidy profile results. In contrast, deviations from the protocol gave very broad histograms. With the protocol, similar results were found for imprints and cytocentrifuge samples. Duplicate "blind" assessments were highly reproducible. DNA indices measured by FCM and ICM (on cytocentrifuge samples) correlated highly as well.

CONCLUSION

With the protocol used, DNA ploidy profiles obtained by ICM from fresh imprints and cytocentrifuge disaggregate cell preparations were comparable.

Confocal 3-dimensional DNA image cytometry in thick tissue sections

We present a three-dimensional confocal DNA image cytometry (3-D CICM) method for analysis of DNA content in 30-40-microns-thick sections of routinely processed paraffin-embedded specimens. A comparison of DNA ploidy profiles obtained by 3-D CICM and conventional DNA image cytometry (ICM) on tissue sections sections showed significantly higher numbers of cells with high DNA content in DNA histograms by 3-D CICM. As estimated by 3-D CICM, the size of nuclei frequently exceeded the thickness of tissue sections used in conventional ICM, which suggested that many nuclei measured by this technique may be incomplete. This artifact was excluded in 3-D CICM by automatic rejection of cut nuclear profiles. This and the favorable ratio of tissue thickness to nuclear size in 3-D CICM permitted the DNA quantitation even in large cells with highly increased DNA ploidy values such as megakaryocytes and Reed-Sternberg cells of Hodgkin's disease. Additionally, 3D-CICM allowed evaluation or morphometric parameters and 3-dimensional reconstruction of studied cells.

Quality control of proliferation marker (MIB-1) in image analysis systems utilizing cell culture-based control materials

Standard controls for quality control of cell growth (proliferation) assays in image analysis systems are not currently available. The authors have developed a system of controls, based on cultured and harvested human cell lines, that can mimic tissue sections. These controls help ensure quality control for the entire cell proliferation analysis process, from the initial cutting of the paraffin block through fixation, immunohistochemical staining, and interactive image analysis. The use of cell line controls is advantageous because of the greater cell population homogeneity and the volume of uniform slides that are obtainable, as opposed to the use of a heterogeneous tissue sample. This system provides an excellent means of evaluating the day-to-day performance of cell proliferation analysis and may also be adapted for use as a method of multi-institutional proficiency testing.

Atypical cells in effusions: diagnostic value of cell image analysis combined with immunocytochemistry

The reliable identification of tumor cells in populations of atypical cells occurring in body cavity effusions is a well-known diagnostic problem. In order to improve tumor cell detection and to predict disease progression, we developed a cell scoring strategy based on a combination of DNA cytophotometry and immunocytochemistry. For this purpose, morphologically atypical cells obtained from 33 effusion samples were submitted to DNA content analysis and tested for Ber-EP4 immunoreactivity. It turned out that elevated DNA content alone has a low specificity (true negative ratio) and sensitivity (true positive ratio) in predicting disease outcome, whereas Ber-EP4 immunoreactivity alone has a high specificity (100%) but a low sensitivity (56%). In contrast, the use of a scoring system combining the two techniques and relating scores to the previous disease state and the cytomorphology of the atypical cells results in highly specific and sensitive prediction of the disease outcome. We therefore suggest that this approach is a valuable tool for reliably identifying tumor cells in effusions containing populations of cytologically suspect cells.

DNA ploidy and MYC DNA amplification in ovarian carcinomas. Correlation with p53 and bcl-2 expression, proliferative activity and prognosis

There is increasing evidence that DNA ploidy is a prognostic factor in ovarian carcinomas, but it is uncertain whether MYC DNA amplification is an epiphenomenon of DNA nondiploidy or a distinct biological change with an impact on the clinical course of the disease. To clarify these issues we analysed DNA ploidy by flow and image cytometry and MYC copy number by polymerase chain reaction in archival material from ovarian carcinomas with known follow up. The results were compared with proliferative activity (Ki67 index) and p53 and bcl-2 expression. DNA cytometry revealed nondiploidy in 84 of 144 cases (58.3%). Nondiploidy was statistically significantly correlated with histological tumour type, histological grade, Ki67 index > 10%, FIGO stage, presence of residual tumour after debulking surgery and adverse postoperative outcome. Furthermore, DNA nondiploidy was associated with p53 accumulation. We found that 84.9% of the p53-positive cases were nondiploid. This points to the paramount importance of wild type p53 for the maintenance of genome integrity in this tumour type. MYC DNA amplification was seen in 33.8% (26/77 cases) of ovarian carcinoma. There was no correlation between MYC DNA amplification and histological tumour type, histological grade, FIGO stage, DNA ploidy, proliferative activity or prognosis. However, when p53 and bcl-2 expression was taken into account, a statistically significant correlation between gene alteration or expression patterns and histological tumour type was revealed. The group of mucinous carcinomas demonstrated both MYC DNA amplification and strong bcl-2 expression in 50% and contained the largest fraction of cases without aberration (37.5%). Endometrioid carcinomas were characterized by strong bcl-2 expression in 85%, whereas serous and undifferentiated carcinomas predominantly exhibited p53 alterations, frequently accompanied by bcl-2 overexpression or MYC DNA amplification. Thus, in interaction with other genes MYC DNA amplification may play a role in the determination of the varying differentiation patterns of ovarian carcinomas.

Effects of topoisomerase I inhibition on the expression of topoisomerase II alpha measured with fluorescence image cytometry

A fluorescence image cytometry technique was developed to measure the effects of topotecan, a topoisomerase I inhibitor, on the nuclear expression of topoisomerase II alpha in a series of patients with refractory or relapsed acute myeloid leukemia (AML). We used a commercially available affinity-purified rabbit polyclonal antibody and a fluorescein-conjugated secondary antibody. By using DAPI as a DNA counterstain and dual wavelength excitation, it was possible to measure enzyme expression in the cell nucleus, and to examine its cell cycle phase distribution. In human acute leukemia cell lines, topoisomerase II alpha expression was greatest in late S and G2 phases, but in leukemia patient samples the enzyme expression appeared to be much less cell cycle dependent. There was considerable interpatient variation in the effects of topotecan on topoisomerase II alpha expression in the leukemia patients, with a threefold increase in the median value after 48 h followed by a decline to pretreatment levels after 5 days of treatment with the topoisomerase I inhibitor. Although these findings should be treated with caution because of the small number of cases studied, they support the prediction that topoisomerase I inhibitors might be capable of increasing sensitivity to topoisomerase II active drugs such as anthracyclines and epipodophyllotoxins by upregulating topoisomerase II expression. They also illustrate the potential value of fluorescence image cytometry for making sequential measurements of the effects of drug resistance modulating agents in cancer patients.

In human hepatocellular carcinoma cells the total membrane surface area of each major organelle is a particular allometric function of the cytoplasmic volume. A morphometric study

Morphometric analysis by both light and electron microscopy was performed in cells from five cases of human, hepatocellular carcinoma (HCC) and in three control cases. In each case, three fragments were examined individually and the following morphometric parameters evaluated: a) nuclear, cytoplasmic and cell volumes; b) volume density and absolute volume of the rough ER, smooth ER, mitochondria, Golgi apparatus, peroxisomes, dense bodies and cytoplasmic matrix; c) surface density, surface/ volume ratio, and total surface area of rough ER, smooth ER and outer mitochondrial membranes. The parameters obtained from HCC cases showed ample scatter of data, all control values lying within the interval between the extreme values for the various parameters. Both the original and the logarithmically transformed data on volume and total membrane surface area of organelles (y) and of the cytoplasmic volume (x) were regressed using first degree regression equations. The original values for volume and total surface area of rough ER, total ER and mitochondria were linearly related to the corresponding values for cytoplasmic volume. The allometric analysis carried out with the logarithms also revealed significant regressions between cytoplasmic volume and smooth ER parameters not detectable when using the original x and y values. It showed, in addition, that in progressively larger cytoplasmic volumes, the cisternae of both rough and smooth ER tend to appear more compacted and a higher portion of the total ER membrane tends to be constituted of smooth ER. Within the wide range of variation in cytoplasmic volume of the HCC cells, the volume and total surface of the organelles do not vary randomly. These data indicate that in the small, normal-sized and large tumoral cells the mechanisms responsible for the cytoplasmic volume and for the corresponding total volume and membrane surface area of each major organelle are interdependent. Such an interdependence gives no support to ideas implying that the variation in size of cancer cells, an element of pleomorphism, would result of anarchical intracellular synthetic and/or degradative conditions.

Clinical importance and prognostic value of the image-DNA-cytometry for patients with gastric cancer

Image-DNA-cytometric analysis was performed retrospectively on air-dried touch preparations from gastric carcinoma from 122 patients who underwent surgery from 1989-1994. The results of DNA cytometry were compared with conventional histologic staging and grading. Aneuploidy could be verified in 81 cases (66.3%), and a significant correlation was demonstrated between aneuploidy and a high incidence of lymph node metastases, whereas no correlation with tumor type and histologic grade was detected. Thus DNA ploidy may by useful in predicting prognosis in gastric cancer.

Three-dimensional confocal laser scanning DNA ploidy cytometry in thick histological sections

DNA ploidy measurement by flow (FCM) or image cytometry (ICM) of single cell suspensions of solid tumour has prognostic value, but it would be a definite advantage if the assessment could be done on histological sections. However, this is usually not possible by means of standard ICM, due to the capping of nuclei in thin sections, or overlap in thick sections. Three-dimensional (3D) microscopy by means of confocal laser scanning microscopy (CLSM) could solve this problem in theory but the results published so far are not very satisfactory. A new method has been developed in which the DNA content of haploid (human testis spermatozoa), diploid, tetraploid, octaploid (human and rat liver and human spermatogonia), and near-triploid (human breast cancer) nuclei stained with YOYO-1 iodide has been measured by a newly developed 3D image cytometry method (3DICM) in 20 microns thick histological sections. YOYO-1 iodide is a new highly sensitive, specific, stoichiometric, and stable fluorescent dye for DNA. DNA ploidy of a breast cancer which was near-triploid with FCM and ICM was also assessed with 3DICM in a tissue section adjacent to the section used for FCM and ICM and the results were compared. The integrated 3DICM fluorescence intensity showed good linearity (r = 0.99) with the real DNA content of all nuclei analysed. In human tissue, the coefficient of variation of 3DICM for haploid (n = 12), diploid (n = 63), triploid (n = 13), tetraploid (n = 12), and octaploid (n = 3) ploidy distributions was 5.1, 6.6, 4.2, 4.0, and 0.6 per cent, respectively (n = the number of nuclei). For the rat liver, the CV of the diploid (n = 21), tetraploid (n = 31), and octaploid (n = 3) peaks was 6.7, 4.8, and 1.6 per cent, respectively. Repeated "blind' measurements of nuclei with different DNA indices showed excellent reproducibility between different observers (r = 0.98). It is concluded that the 3DICM method used is accurate, reproducible, and clinically feasible in thick histological sections. This is especially important in small lesions, or if the results of DNA ploidy measurement of single cell suspensions (by FCM) or imprints (by ICM) are inadequate.

DNA image cytometry on sections as compared with image cytometry on smears and flow cytometry in melanoma

DNA measurements of 130 melanomas were carried out by flow cytometry (FCM) and image cytometry (ICM). ICM was applied to cytological preparations of fresh material (cICM) and to sections of formalin-fixed paraffin embedded tissue (sICM). The DNA ploidy, the DNA index of G0/G1 peaks (DI), and the proliferation index (PI) were used to compare all the methods. The following parameters reflecting malignancy were calculated only from ICM histograms: the 5c exceeding rate (5cER) and the malignancy grade (MG). In cases found to be DNA aneuploid by FCM, the PI values (FCM versus cICM) and the DIs (between all methods) showed a high correlation, and the concordance in relation to the DNA ploidy status was 96% (FCM versus cICM) and 94% (FCM versus sICM). However, we ascertained essential differences between FCM and ICM in melanomas classified as DNA diploid by FCM. The concordance in DNA ploidy was only 66% (FCM versus cICM) and 64% (FCM versus sICM). In contrast, cICM and sICM yielded similar results in most cases. With the exception of the near diploid range, ICM is superior to FCM in detecting DNA aneuploidy. In particular, DNA tetraploid stem lines can easily be overlooked by FCM. Therefore, DNA measurements of tumours judged to be DNA diploid by FCM must be verified by ICM. ICM on sections proved to be applicable and yielded reliable results provided that a suitable thickness was used, and the measuring of sectioned and overlapping nuclei was largely avoided by careful focusing in either direction.

Comparison of different mathematical algorithms to correct DNA-histograms obtained by measurements on thin liver tissue sections

Several mathematical algorithms have been published so far to correct histograms obtained from DNA-measurements on thin sections. All these methods require knowledge of the thickness of the section currently under investigation. As the problem of determining the actual section thickness has recently been solved, the different methods for recalculation can be compared under strictly controlled conditions. For histogram recalculation, the algorithms published by Bins, McCready, Bacus, Rigaut and Haroske are applied on a series of rat liver tissue sections with increasing section thicknesses. The histograms are compared to that of measurements of a single cell preparation of the same tissue block. Our results suggest that DNA-measurements on thin paraffin sections are possible if the actual section thickness is known. Correct ploidy equivalents throughout the whole spectrum of section thicknesses investigated are calculated using the formulas of McCready and Haroske. This is also the case using the correction method of Bins in very thin sections. The algorithms of Bacus and Rigaut do not lead to reliable results. Recalculated histograms show some differences to those obtained by DNA-measurements of single cell preparations. The histograms show broader CVs of the peaks and lymphocytes as internal ploidy standard are not useful in the majority of the algorithms. Nevertheless the promising results of this study legitimize efforts to establish standards for the interpretation of recalculated histograms in the near future. The broader applicability to tumour sections remains unproven. Therefore, further investigation are necessary using more heterogeneous tissue including aneuploid tumour cell populations.

A strategy for multiple immunophenotyping by image cytometry: model studies using latex microbeads labeled with seven streptavidin-bound fluorochromes

Multiple immunophenotyping is aimed at identifying several cell populations in a single labeling procedure by their ability to bind combinations of specific labeled antibodies. The present work demonstrates the simultaneous discrimination by using image cytometry of aminomethylcoumarin acetate (AMCA), Lucifer yellow (LY), fluorescein isothiocyanate (FITC), R-phycoerythrin (PE), PE-Texas red tandem (Red613), peridinin-chlorophyll protein (PerCP), and allophycocyanin (APC), which were all bound to latex beads as streptavidin-conjugated fluorochromes. This has been the result of a step-by-step optimization of the several factors affecting the sensitivity and specificity of multiple immunofluorescence analysis. First, 14 streptavidin-conjugated fluorochromes were evaluated by using spectrofluorometry. A primary selection was then made of ten spectrally separable dyes that could be evaluated by using image cytometry. These dyes were bound to latex particles, and specific filter combinations were assembled to minimize crosstalk between fluorophores while preserving sufficient fluorescence intensity and counting statistics. Potential probe associations were then assessed by measuring the emissions of all fluorochromes that were detected by each filter combination. The resulting crosstalk matrix served as the basic tool both for final selection of the optimal filter combination and for dye set (composed, in this case, of the seven fluorochromes described above) and for mathematical correction of residual spectral overlap. Next, an image cytometry system was adapted to collect seven images of matched brightness with the selected combination of excitation/emission filters and dichroic mirrors. Finally, seven-parameter synthetic images were generated by digital image processing.

Photothermal image cytometry of human neutrophils

Photothermal imaging, when applied to the study of living cells, provides morphofunctional information about cell populations. In technical terms, the method is complementary to optical microscopy. The photothermal method was used for cell imaging and quantitative studies. We have developed a dual-laser technique and have constructed an automated image cytometer based on an optical microscope. The theory of the method, the technical equipment, and the computer software are described. With our technique, three types of cell images can be acquired and analyzed: optical, pulsed laser time-domain, and photothermal. The parameters of each kind of image are considered in detail. Preliminary results of studies on living human neutrophils are presented.

Cell cycle expression of estrogen receptors determined by image analysis on human breast cancer cells in vitro and in vivo

We have investigated, by image analysis, the cell cycle expression of estrogen receptors (ER) on MCF-7 cell line and on MCF-7 xenografts. The results demonstrate, in vitro as well as in vivo, an increase of ER concentration during the G0/G1-phase, followed by a decrease during the S-phase until the late S-phase where a rapid increase was noted. These results confirm that estrogens are involved in the DNA synthesis since ER is expressed in vivo at a maximal level in the late G1. In presence of saturating concentrations of 17 beta-estradiol, the mean ER concentration in G0/G1 phase is significantly decreased compared with the control cells cultured in estrogen-deprived medium. This indicates that 17 beta-estradiol down-regulates ER preferentially in the G0/G1 phase. These data suggest that ER in S and G2/M phases is unable to interact with its ligand. Consequently, estrogens may have no effects on the entry of cells in mitosis. Finally, after long-term tamoxifen treatment of MCF-7 xenografts, a tamoxifen-resistant tumor was developed which was characterized by a change in the profile of ER concentration during the G0/G1 phase. In conclusion, it is possible that the differences in cell cycle distribution of ER could be correlated with different phenotypes of breast cancer and also with different clinical phases of tumoral evolution. However, it remains to be known what is the clinical significance of the ER cell cycle expression in relation to tumor aggressiveness and survival.

Image cytometry and topographical analysis of proliferation of endothelial cells in vitro during Bartonella (Rochalimaea) infection

Bartonella quintana and Bartonella henselae are clinically associated with proliferative neovascular lesions. The effect of Bartonella infection on human endothelial cells was evaluated in vitro by quantitative image analysis. Particular emphasis is placed on reporting the methodologies employed. Human umbilical vein endothelial cells were infected in vitro with the two Bartonella species. Cell proliferation (cell density), cell morphology (cell surface, form and elongation factors) and spatial reorganization (global topographical analysis and hierarchical cluster detection) were monitored over a 3-day period of infection. Firstly, infection stimulated endothelial cell proliferation. Secondly, infection induced obvious morphological changes; infected cells became larger, more elongated and spindle-shaped. Cytoskeletal reorganization was confirmed by staining of F actin. Thirdly, infection altered the spatial organization of cells within the monolayer; this could not have been due solely to the morphological modifications they experienced. This model demonstrates that Bartonella infection provoked endothelial cell proliferation, topographical rearrangements and morphological changes because of modifications of the cytoskeleton. These experimental findings provide a physiopathological explanation to the abnormal angiogenesis observed in bacillary angiomatosis.

Immunophenotypic analysis of hematologic malignancy by laser scanning cytometry

The authors tested a newly-developed computerized laser scanning cytometer (LSC) as a means of performing immunophenotypic analysis of hematologic specimens within their community hospital. Results were compared on a case-by-case basis with parallel flow cytometric and immunohistochemical data. A total of 71 specimens analyzed include 22 excised lymph nodes or other tissue biopsies, 18 peripheral bloods, 17 bone marrow aspirates, 7 body fluids, and 7 fine-needle aspiration biopsies of lymphoid tissue. The LSC proved to be a useful instrument capable of generating simultaneous two-color immunofluorescent data directly analogous to that obtained via conventional flow cytometry. However, laser scanning cytometric analysis provides advantages over flow cytometric analysis, because the LSC measures cells on a slide rather than in a fluid stream. Specifically, cells can be microscopically examined at any time--before, during, or after automated immunofluorescent analysis. In addition, specimen preparation techniques are less restricted and more cost efficient. Lastly, even extremely small and/or hypocellular specimens (such as body fluids and fine-needle aspiration biopsies) can be successfully analyzed.

Methods in quantitative image analysis

The main steps of image analysis are image capturing, image storage (compression), correcting imaging defects (e.g. non-uniform illumination, electronic-noise, glare effect), image enhancement, segmentation of objects in the image and image measurements. Digitisation is made by a camera. The most modern types include a frame-grabber, converting the analog-to-digital signal into digital (numerical) information. The numerical information consists of the grey values describing the brightness of every point within the image, named a pixel. The information is stored in bits. Eight bits are summarised in one byte. Therefore, grey values can have a value between 0 and 256 (2(8)). The human eye seems to be quite content with a display of 5-bit images (corresponding to 64 different grey values). In a digitised image, the pixel grey values can vary within regions that are uniform in the original scene: the image is noisy. The noise is mainly manifested in the background of the image. For an optimal discrimination between different objects or features in an image, uniformity of illumination in the whole image is required. These defects can be minimised by shading correction [subtraction of a background (white) image from the original image, pixel per pixel, or division of the original image by the background image]. The brightness of an image represented by its grey values can be analysed for every single pixel or for a group of pixels. The most frequently used pixel-based image descriptors are optical density, integrated optical density, the histogram of the grey values, mean grey value and entropy. The distribution of the grey values existing within an image is one of the most important characteristics of the image. However, the histogram gives no information about the texture of the image. The simplest way to improve the contrast of an image is to expand the brightness scale by spreading the histogram out to the full available range. Rules for transforming the grey value histogram of an existing image (input image) into a new grey value histogram (output image) are most quickly handled by a look-up table (LUT). The histogram of an image can be influenced by gain, offset and gamma of the camera. Gain defines the voltage range, offset defines the reference voltage and gamma the slope of the regression line between the light intensity and the voltage of the camera. A very important descriptor of neighbourhood relations in an image is the co-occurrence matrix. The distance between the pixels (original pixel and its neighbouring pixel) can influence the various parameters calculated from the co-occurrence matrix. The main goals of image enhancement are elimination of surface roughness in an image (smoothing), correction of defects (e.g. noise), extraction of edges, identification of points, strengthening texture elements and improving contrast. In enhancement, two types of operations can be distinguished: pixel-based (point operations) and neighbourhood-based (matrix operations). The most important pixel-based operations are linear stretching of grey values, application of pre-stored LUTs and histogram equalisation. The neighbourhood-based operations work with so-called filters. These are organising elements with an original or initial point in their centre. Filters can be used to accentuate or to suppress specific structures within the image. Filters can work either in the spatial or in the frequency domain. The method used for analysing alterations of grey value intensities in the frequency domain is the Hartley transform. Filter operations in the spatial domain can be based on averaging or ranking the grey values occurring in the organising element. The most important filters, which are usually applied, are the Gaussian filter and the Laplace filter (both averaging filters), and the median filter, the top hat filter and the range operator (all ranking filters). Segmentation of objects is traditionally based on threshold grey values. (AB

Assessment of regional cytochrome P450 activities in rat liver slices using resorufin substrates and fluorescence confocal laser cytometry

Characterizing constitutive activities and inducibility of various cytochrome P450 isozymes is important for elucidating species and individual differences in susceptibility to many toxicants. Although expression of certain P450s has been studied in homogenized tissues, the ability to assess functional enzyme activity without tissue disruption would further our understanding of interactive factors that modulate P450 activities. We used precision-cut, viable rat liver slices and confocal laser cytometry to determine the regional enzyme activities of P450 isozymes in situ. Livers from control and beta-naphthoflavone (beta NF)-treated rats were sectioned with a Krumdieck tissue slicer into 250-microns thick sections. A slice perfusion chamber that mounts on the cytometer stage was developed to allow for successive measurement of region-specific P450-dependent O-dealkylation of 7-ethoxy-, 7-pentoxy-, and 7-benzyloxyresorufin (EROD, PROD, and BROD activity, respectively) in the same liver slice. Images of the accumulated fluorescent resorufin product within the tissue were acquired using a confocal laser cytometer in confocal mode. As expected, slices isolated from beta NF-treated rats showed high levels of centrilobular EROD activity compared to slices from control rats, whereas PROD and BROD activities remained at control levels. These techniques should allow for the accurate quantification of regional and cell-specific P450 enzyme activity and, with subsequent analysis of the same slice, the ability to correlate specific P450 mRNAs or other factors with enzymatic activity. Moreover, these techniques should be amenable to examination of similar phenomena in other tissues such as lung and kidney, where marked heterogeneity in cellular P450 expression patterns is also known to occur.

Cytometry with DAPI-stained tumor imprints. A reliable tool for improved intraoperative analysis of lung neoplasms

OBJECTIVE

To determine the feasibility and potential use of application of 4'-6-diamidino-2-phenylindole dihydrochloride-stained imprints taken from primary lung carcinomas for intraoperative frozen section service.

STUDY DESIGN

Tumor imprints taken from intraoperative sections were stained with the fluorochrome DAPI, and the integrated nuclear fluorescence (INF) was measured using an automated image analyzing system. The 40 cases studied comprised all major lung carcinoma types as well as intrapulmonary metastases. The measured features included percentage of tumor cells > 3C and > 5C, S-phase-related tumor cell fraction, 2 CV standard deviation and INF entropy.

RESULTS

Test series with imprints of tumor-free lymph nodes revealed a minimum staining time of five minutes at 40 degrees C that led to reproducible staining intensities and size of the nuclear area. No significant differences were apparent in comparison to concomitantly performed measurements of the integrated optical density of "classical" Feulgen-stained specimens for each case. The fluorometric results could be submitted to the surgeon with 15-18 minutes (20 maximum).

CONCLUSION

The data indicate that cytometric DNA measurements can be performed intraoperatively with reliable quality within time schedules acceptable for intraoperative surgical strategies.

Identification of F-reticulocytes by two-stage fluorescence image cytometry

Precise determination of reticulocytes (young red blood cells) containing fetal hemoglobin (Hb F), F-reticulocytes, is important for assessment of the efficacy of drugs such as hydroxyurea and butyrate in elevating the levels of Hb F in patients with sickle cell disease (SCD) and beta-thalassemia. We developed a reliable and easily applicable method for determining F-reticulocytes using fluorescence image cytometry. Reticulocytes were first identified by preparing a monolayer smear of blood stained by acridine orange. Images of reticulocytes and of all cells were obtained for a selected area on the smear. After removing the acridine orange, cells containing Hb F (F-cells) in the same area were then identified by immunofluorescence. Using images of F-cells, reticulocytes, and all cells for the same fields it was possible to identify F-reticulocytes. To assess the validity of our two-stage staining method, we compared our results with those obtained by traditional methods. There was significant correlation of our method with the conventional immunofluorescence staining method for F-cells (r2 = 0.99; slope = 0.99) and with the accepted brilliant cresyl blue method for reticulocytes (r2 = 0.97; slope = 0.96). Heretofore, the ability to determine F-reticulocyte levels has been limited to a small number of laboratories possessing special equipment and techniques. The method presented here should be of great interest to many basic science and clinical investigators involved in studies evaluating synthesis of Hb F.

Volumetric capillary cytometry: a new method for absolute cell enumeration

This article describes a novel cytometry system that is capable of imaging, identifying, and counting populations of fluorescently labeled cells within a precisely known volume of unlysed whole blood contained in a precision capillary. The whole blood is reacted with dye-labeled antibodies directed at cell surface antigens. A red excitation source (helium-neon laser) is used to minimize auto-fluorescence and interference by blood components. A two-color fluorescence detection system is described, using the dyes Cy5 and Cy5.5, allowing the simultaneous detection and quantitation of two different types of surface antigens. Image processing techniques are used to identify cellular events in the image and to quantitate the fluorescent intensities in both color planes. The utility of this technique is demonstrated by performing absolute CD3+/CD4+ lymphocyte counts in whole blood, with minimal sample preparation, and with the blood sample completely contained in a disposable cartridge. A correlation study comparing this technique to flow cytometry gives a correlation coefficient of 0.98. In addition, the ability of this system to monitor reaction kinetics is demonstrated with a measurement of the rate of association of the anti-CD4 antibody onto the surface of CD4+ lymphocytes.

DNA ploidy analysis by laser scanning cytometry (LSC) in colorectal cancers and comparison with flow cytometry

We evaluated laser scanning cytometry (LSC) by comparing nuclear DNA ploidy determined by LSC and by flow cytometry (FCM) in 77 samples of human colorectal cancer from 48 patients. Both methods revealed an aneuploid peak in 30 (62.5%) of the cases, although two samples that were aneuploid by LSC were diploid by FCM and two others were diploid by LSC and aneuploid by FCM. The concordance rate for nuclear DNA ploidy was 91.7% in the 48 patients and 87.0% for the 77 samples. The DNA index was also highly correlated between two methods (r2 = 0.97, P < 0.001). We concluded that LSC provides DNA histograms equivalent to FCM for surgical specimens and has potential clinical application in pathology.

Cell cycle expression of steroid receptors determined by image analysis on human breast cancer cell line: a hypothesis on the effects of antiestrogens

Tamoxifen is the widespread anti-hormonal compound used for the treatment of human breast cancer. It is admitted that its effects are mediated via estrogen receptors (ER) but the molecular basis of its activity has yet to be clearly defined. In this work, we have developed a new image cytometry procedure in order to clarify the interactions between steroid receptors and tamoxifen at the cell cycle kinetic level. On untreated cells, an increase of ER level and a decrease of progesterone receptor (PR) level during the G0/G1 phase were demonstrated. Then, the ER and PR levels fell during the S-phase until the beginning of G2/M phase, where an increase was observed, especially for PR. These results suggest that ER is synthesized preferentially during the G0/G1 transition and PR during the S/G2 transition. After short-term tamoxifen treatment an augmentation of ER level was observed which was not dose-dependent, suggesting an increase in receptor translation rather than an augmentation of ER synthesis. PR level declined in the majority of the population leading to a selection of a subset of proliferating PR negative cells after treatment. These data demonstrate that the synthesis of steroid receptors is linked with the progression of cells through the cell cycle and indicate that tamoxifen blocks MCF-7 cells in G1 via its interactions with ER. Our multifluorescence imaging procedure appears to provide a rapid and quantitative approach which is useful for investigating alterations in steroid receptors after endocrine treatment.